Home

Preliminary Screening System for Ambient Air

1. C 19 entries sumSmallLoc1 countLoc Data 1 sumLargeLoc1 countLoc1 Data 1 150 high Write to separate files Location 1 through 5 Location 1 if locationNumber 1 outputLoc1 lt lt setw 6 lt lt left lt lt LoclIndex lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 16 lt lt left lt lt concatDateTime lt lt lt lt setw 5 lt lt left lt lt nodeldentifier lt lt lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 12 lt lt left lt lt humidity lt lt lt lt setw 11 lt lt left lt lt condition lt lt lt lt setw 12 lt lt left lt lt smallCountInt lt lt lt lt setw 12 lt lt left lt lt largeCountInt lt lt lt lt setw 13 lt lt left lt lt conc25 gaT lt lt setw 12 lt lt left lt lt conc10 lt n lt lt setw 7 lt lt left lt lt message lt lt endl LoclIndex Increase to next row value count data if Locl Index gt 2 if prevDay1 dayID double smallAverageLoc1 double largeAverageLocl Warning messages string messageLoc 1 Warn 3 both PM2 5 and PM10 over thresholds if smallAverageLocl gt 35 amp amp largeAverageLoc 1 gt messageLocl Warning PM 2 5 and PM 10 too Warn 1 PM2 5 over 35ug m 3 elseif smallAverageLocl gt 35 messageLocl Warning PM 2 5 too high Warn 2
2. amp BeginMonth amp amp BeginDay Sheets Add Name DateTitle Sheets DateTitle Select Property used to capture information from myphl17 weather With Sheets DateTitle QueryTables Add Connection _ URL http weather myphl17 com auto wb17 history airport amp LocationID amp BeginDate amp DailyHistory html _ Destination Range A 1 Name DailyHistory FieldNames True RowNumbers False FillAdjacentFormulas False PreserveFormatting True RefreshOnFileOpen False BackgroundQuery True RefreshStyle xlInsertDeleteCells SavePassword False SaveData True AdjustColumn Width True RefreshPeriod 0 WebSelectionType xlSpecifiedTables WebFormatting x1WebFormattingNone WebTables 6 WebPreFormattedTextToColumns True WebConsecutiveDelimitersAsOne True WebSingleBlockTextImport False WebDisableDateRecognition False WebDisableRedirections False Refresh BackgroundQuery False End With Format time to match with algorithm program Columns A A Select Selection NumberFormat h mm Save copy of weather data to bring into algorithm program Application DisplayAlerts False ActiveSheet SaveAs Filename C Documents and Settings amp UserName amp Desktop Weather Output Output from Import Weather Data Weather amp DateTitle amp csv FileFormat xICSV _ CreateBackup False Application DisplayAlerts True C 2 BeginDay BeginDay 1
3. 1 Exclude lines with errors in size These contain two identifiers in the same line but not two sets of data identifier contain such as H int identifierFrontLoc line find int identifierBackLoc line rfind if identifierFrontLoc identifierBackLoc Exclude lines with no identifier included in line This occurs when the data is not fully received This usually occurs at the last line of the file if identifierFrontLoc gt 0 All important lines are now included from the RealTerm Output file The useful information from each line now will be extracted EEEE EE EEEE EEE EEEE EEEE EEEE E EEEE CIGS ICIS ICSC ICICI CIRCE Part 3 Extraction of useful data from RealTerm Output Including Time stamp node identifierm small count large count PRR R SCRE EEEE EEE EEEE EEE EEE EEEE EE E ISIS ICSC CIC IIE EEEE EEEE EEEE EEEE E EEE E E EE EEEE E EE E EEE EEE EEEE E EEEE EEEE EEEE CII Part 3a Extraction of Time Stamp Reformat to include day month time PE gt A a e ee hee he ee ake a ee ae ae ake ake ae a a ake ae he 2 ake ake ake ake afe ake e he ae ae a fe ake 2 a ak he 2 ake ae fe ake 2 fe a ake 2 ee 2 fe afe fe ake afe he 2 ae ake fe ale a ak 2 ake 2k Extract Unix Time Stamp find first int firstMarkerLoc line find string timeStamp line substr 0 firstMarkerLoc Convert time stamp to Int int timeStampInt atoi timeStamp c_str Ad
4. Comma 3 Location int comma3Loc weatherLine find comma2Loc 1 comma Loc 1 Find Humidity Reading string humidityString weatherLine substr comma3Loc 1 4 int percentLoc humidityString find string humidityValue humidityString substr 0 percentLoc Convert humidity to an integer Particle Count Convert from string to integer to double int humidityInt atoi humidityValue c_str humidity atof humidityValue c_str Comma 4 Location int comma4Loc weatherLine find Comma 5 Location int commaSLoc weatherLine find Comma 6 Location int comma6Loc weatherLine find comma5Loc 1 Comma 7 Location int comma7Loc weatherLine find comma6Loc 1 Comma 8 Location int comma8Loc weatherLine find Comma 9 Location int comma9Loc weatherLine find Comma 10 Location int comma10Loc weatherLine find comma9Loc 1 comma3Loc 1 comma4Loc 1 comma7Loc 1 comma8Loc 1 C 15 Find Condition string condString weatherLine substr commal0Loc 1 6 int condLoc condString find string condition condString substr 0 condLoc Apply separate factor for rain if condition Rain Rain Table Improvement Factor double rain50 2 5 50 59 double rain60 2 2 60 69 double rain70 2 0 70 79 double rain80 1 4 80 89 double rain90 0 8 90 100 Humidity between 50 59 if humidityInt
5. Wireless Module Wireless Module Air Sensor Air Sensor Wireless Module Air Sensor Figure 1 Overview of System Data Flow and Actual Node Configuration National Ambient Air Quality Standards Primary Standards Secondary Standards Pollutant Averaging Time Averaging Time Particulate 3 Same as Primary Matter PM19 Particulate 15 0 pg m a Same as Primary Matter PM2 5 Same as Primary Table 1 EPA National Ambient Air Quality Standard Data retrieved from a filter takes approximately two weeks to process due to the labor involved For this reason the AMS has installed continuous particulate matter monitors Met One Instruments Model BAM 1020 at some of its testing sites While these devices can give an instantaneous PM concentration they are labeled only Federal Equivalent Method certified by the EPA meaning that they can not be used as the final judge of air quality against the EPA standard Furthermore the devices are expensive costing 14 300 each 8 9 The preliminary screening system that has been developed is a cost effective alternative to the current method of evaluating the amount of particulate matter in ambient air When deployed in a given neighborhood it will determine whether or not the cost and time of the more precise EPA funded tests should be invested This project is being performed to the specifications of the Clean Air Council of Philadelphia CAC The CAC ha
6. aj bole tT riot iS mi i RE Sote n SY Figure 21 Southeast Philadelphia Deployed Node Houston Community Center E 1 Figure 22 Senior Design Team Mentoring SLA Students Appendix F EPICS Mentoring Hi Hs il Figure 23 Senior Design Team Performs a Sensor Height Test with SLA Students F 1 Appendix G System Hardware Technical Discussion Sensors Sensor Decision Process Several off the shelf components were evaluated for the actual capturing of ambient air The options were considered based on several characteristics including sensitivity precision price and power consumption As previously stated the EPA utilizes a filtering system to test for particulate matter thus this alternative was explored In addition lower cost alternatives were also discussed including household air quality detectors similar to those used by design team ECE 17 in 2007 2008 However because these sensors resulted in little success 11 a different cost effective solution was explored the laser particle counter Heralded for its mobility and ability to measure particle concentrations in short time intervals in addition to its low cost 12 this device was seen as a very attractive solution for the proposed design Upon researching the alternatives it was determined that the Dylos DC1100 Pro laser particle counter would be most appropriate for the solution at hand Though it is less precise than the
7. Conc PM 2 5 lt lt setw 13 lt lt left lt lt Conc PM 10 lt lt setw 8 lt lt left lt lt Message lt lt endl Location 5 ofstream outputLoc5 outputLoc5 open C Documents and Settings Anthony Desktop C Results Location 5 Data txt outputLoc5 Title lt lt setw 16 lt lt left lt lt Location 5 Data lt lt endl lt lt endl Header Header lt lt setw 6 lt lt left lt lt Index lt lt setw 6 lt lt left lt lt WDay lt lt setw 17 lt lt left lt lt Date Time lt lt setw 6 lt lt left lt lt Node lt lt setw 36 lt lt left lt lt Node Location lt lt setw 13 lt lt left lt lt Humidity lt lt setw 12 lt lt left lt lt Conditions lt lt setw 13 lt lt left lt lt Small Count lt lt setw 13 lt lt left lt lt Large Count lt lt setw 14 lt lt left lt lt Conc PM 2 5 lt lt setw 13 lt lt left lt lt Conc PM 10 lt lt setw 8 lt lt left lt lt Message lt lt endl C 7 Daily Averages ofstream averages averages open C Documents and Settings Anthony Desktop C Results Averages txt averages Title lt lt setw 20 lt lt left lt lt Daily Averages All Locations lt lt endl lt lt endl Header lt lt setw 36 lt lt left lt lt Location lt lt setw 6 lt lt left lt lt WDay lt lt setw 6 lt lt left lt lt Y ear lt l
8. PM2 5 3 ug m wom x Conoeniralon soe ane ne eo nena nwo nme nasa nawo woe uono weno wn eo maso waso mass Figure 16 Comparison of AMS Data to ECE 19 Corrected Data Sensor 2 Comparison of Daily Averages Filter Method to ECE 19 Sensor1 and AMS Continuous Data 25 00 e AMS Filter PM 2 5 E 20 00 E AMS Continuous PM gt 2 5 MetOne E 15 00 4 ECE19 Sensor 1 PM2 5 2 10 00 8 S amp 5 00 0 00 4 3 2009 4 5 2009 4 7 2009 4 9 2009 4 11 2009 4 13 2009 4 15 2009 4 17 2009 Figure 17 Comparison of Daily Averages Filter Method to AMS and ECE 19 Continuous Data Filter ECE 19 Sensor AMS MetOne ECE Sensor 1 DifiMetOne Diff 4 5 2009 88 1 47 ECE 19 4 6 2009 4 J 0 55 ECE 19 4 7 2009 i i 1 97 MetOne 4 8 2009 4 i 2 15 MetOne 4 9 2009 is i 1 64 MetOne 4 10 2009 2 52 ECE 19 4 11 2009 i 4 i 4 89 ECE 19 4 12 2009 1 09 MetOne 4 13 2009 0 82 MetOne 4 14 2009 1 22 ECE 19 4 15 2009 0 97 ECE 19 0 22 MetOne Table 5 Comparison of Daily Averages Filter Method to AMS and ECE 19 Continuous Data D 2 Figure 18 Size Comparison of System Node Size Small Box to FEM Monitor Large Box D 3 Figure 19 AMS Filter Weighing System D 4 Appendix E Southeast Philadelphia Deployment a w ae Pa sh au amas 4 EY meee GC GLU L SiS tht D gt We orai ki
9. Parameter View Versions versions Version Download new Remote Configuration PC Settings Range Test Terminal Modem Configuration Modem Parameters and Firmware r Parameter View r Profile Read Write Restore Clear Screen Save Show Defaults Load J Versions I Always update firmware E Modem XBEE PRO Function Set Download new VISIONS Version Modem XBEE PRO Function Set_ 1820 v XBP09 DM XBEE PRO DIGIMESH 900 Sa 1820 v 2 B Mo PWMO duty cycle a f 0 HP Hopping Channel 1 MR Mesh Retries D 7 NH Network Hops Bi 3 NO Network RREG Retries Network Delay Slots Mesh Networking G 134200 SH Serial Number High D 4055BACA SL Serial Number Low B 0 DH Destination Address High f FFFF DL Destination Address Low COORDINATOR NI Node Identifier SC NT Node Discovery Backoff 43 Security 0 EE Encryption Enable D KY AES Encryption Key Serial Interfacing B 3 8D Baud Rate 0 NB Parity Change networking settings COM9 9600 8 N 1 FLOW NONE XBP09 DM B M1 PWM1 duty cycle 20 RP RSSI PWM Timer D 3F7F PR Pull up Resistor Enable Sy AT Command Options f 64 CT AT Command Mode Timeout 38 GT Guard Times 28 CC Command Sequence Character 44 Diagnostic Commands D 1820 VR Firmware Version 1843 HY Hardware Version D 024 XV Supply Voltage D 0 ER Receive Error Count D 0 GD
10. Save Download F J Always update firmware Show Defaults Load versions Modem XBEE PRO Function Set _ Version XBPOSDM XBEE PRO DIGIMESH 900 x 1820 x D MO PWMO duty cycle a B M1 PWM1 duty cycle J D RP RSSI PWM Timer f PR Pull up Resistor Enable 4 AT Command Options B CT AT Command Mode Timeout D GT Guard Times B CC Command Sequence Character Diagnostic Commands VA Firmware Version B HV Hardware Version D 2 Supply Voltage D ER Receive Error Count D GD Receive Good Count D TR Delivery Failure Count sep Commands B 1 S0 Sleep Options fa 4 SM Sleep Mode FDE8 ST Wake Period E DOFC SP Cyclic Sleep Period Configure sleeping parameters COM3 9600 8 N 1 FLOW NONE Figure 8 Sleep and Wake Parameters J 8 E Configuration of System Node Completion of Cyclic Sleep Configuration After the XBee Pro Coordinator module USB interface has been configured with the proper sleep parameters each XBee Pro node module serial interface will need to be programmed individually by utilizing a serial to USB cable 1 Connect the XBee Pro serial interface board via a serial to USB cable 2 Re open the X CTU Software first close all existing instances 3 Go to the Modem Configuration tab and press Read 4 Scroll down to Sleep Commands 5 Set SO Sleep Options to 0 PREFERRED SLEEP COORDINATOR DISABLED 6 Se
11. lt lt setw 11 lt lt left lt lt condition lt lt lt lt setw 12 lt lt left lt lt smallCountInt lt lt lt lt setw 12 lt lt left lt lt largeCountInt lt lt lt lt setw 13 lt lt left lt lt conc25 lt lt lt lt setw 12 lt lt left lt lt concl0 Sy lt lt setw 7 lt lt left lt lt message lt lt endl Loc2Index Increase to next row value count data entries if Loc2Index gt 2 if prevDay2 dayID double smallAverageLoc2 sumSmallLoc2 countLoc2Data 1 double largeAverageLoc2 sumLargeLoc2 countLoc2Data 1 C 21 150 high Rats n x Warning messages string messageLoc2 Warn 3 both PM2 5 and PM10 over thresholds if smallAverageLoc2 gt 35 amp amp largeAverageLoc2 gt messageLoc2 Warning PM 2 5 and PM 10 too Warn PM2 5 over 35ug m 3 elseif smallAverageLoc2 gt 35 messageLoc2 Warning PM 2 5 too high Warn 2 PM10 over 150ug m 3 else if largeAverageLoc2 gt 150 messageLoc2 Warning PM 10 too high No Warning else messageLoc2 averages lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 5 lt lt left lt lt yearID Ken lt lt setw 6 lt lt left lt lt monthID lt lt lt lt setw 4 lt lt left lt lt dayID lt lt lt lt setw 20 lt lt left lt lt smallAverageLoc2
12. lt lt setw 12 lt lt left lt lt smallCountInt lt lt lt lt setw 12 lt lt left lt lt largeCountInt lt lt lt lt setw 13 lt lt left lt lt conc25 LATT lt lt setw 12 lt lt left lt lt conc10 KaU lt lt setw 7 lt lt left lt lt message lt lt endl LocSIndex Increase to next row value count data entries if LocSIndex gt 2 if prevDay5 dayID double smallAverageLoc5 sumSmallLoc5 countLoc5Data 1 C 25 sumLargeLoc5 countLoc5Data 1 150 high lt lt en i double largeAverageLoc5 Warning messages string messageLocsS Warn 3 both PM2 5 and PM10 over thresholds if smallAverageLoc5 gt 35 amp amp largeAverageLoc5 gt messageLoc5 Warning PM 2 5 and PM 10 too Warn 1 PM2 5 over 35ug m 3 elseif smallAverageLoc5 gt 35 messageLoc5 Warning PM 2 5 too high Warn 2 PM10 over 150ug m 3 else if largeAverageLoc5 gt 150 messageLoc5 Warning PM 10 too high No Warning else messageLoc5 averages lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 5 lt lt left lt lt yearID gann lt lt setw 6 lt lt left lt lt monthID lt lt lt lt setw 4 lt lt left lt lt dayID lt lt lt lt setw 20 lt lt left lt lt smallAverageLoc5 lt lt setw 20 lt lt left lt lt largeAverageLoc
13. Light Drizzle 8 F 51 8 F 100 30 05 in 3 0 miles NE 11 5 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 5 0 miles NNE 11 5 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 2 0 miles North 6 9 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 1 8 miles North 5 8 mph N A Light Drizzle 8 F 51 8 F 100 30 04 in 1 5 miles North 6 9 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 04 in 1 5 miles NNE 6 9 mph 0 01 in Light Drizzle 1 F 51 1 F 100 30 04 in 1 2 miles North 5 8 mph 0 01 in Light Drizzle 8 F 51 8 F 100 30 04 in 1 0 miles NNE 11 5 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 2 0 miles NNE 11 5 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 2 0 miles North 6 9 mph 0 00 in Light Drizzle 1 8 F 51 8 F 100 30 05 in 2 0 miles North 5 8 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 1 8 miles North 8 1 mph 0 00 in Light Drizzle Figure 20 Output File from Import Weather Data Program 2 Data Extraction and Algorithm Implementation The Data Extraction and Algorithm Implementation program is the most important portion of the software package This program reads in both the RealTerm output and the J 18 weather data These files are used to effectively perform the particle count to mass conversion algorithm More specifi
14. RealTerm Program and Output File Located on the Desktop 1 RealTerm RealTerm is a serial terminal program that is used to capture the sensor data This program allows the coordinator module to communicate through a COM port and output the received data to a text file Installation of Program 1 Download Realterm at http realterm sourceforge net index html downloads_Download 2 Run the setup file to install Realterm on the same computer where the coordinator is located Configuration of Port 3 Open Realterm 4 On the Port Tab make sure the settings match those of the X CTU software e Port Same port that the USB drivers were installed on Baud Rate 9600 Parity None Bits 8 Flow Control None Stop Bits 1 5 Click Change to apply changes As shown in Figure 15 Display Port Capture Pins Send Echo Port 120 120 2 I2CMisc Misc n Clear Freeze _ Status Baud 9600 Port E Open Spel A Change SE ioe et _ XD 2 Parity Data Bits Stop Bits potters fioi coat TXD 3 a D E Recei 17 E None Shits 1bt C 2bis m SNE aii Char fi ICTS 8 r pa 7 bits Hardware Flow Control I Transmit Xoff Char 19 _ DCD 1 C Mak Shits None C RTSYCTS _ DSR 6 Space C 5bits C DTR DSR C RS485 ts e J Ring 9 G _ BREAK Error Settings only change when you click this j Char Count 0000000 cPs 0 j Port 1 57600 8N1 None Figure 15 RealTerm Port Tab Configu
15. Receive Good Count Bs OTH Deh e Sleep Mode Sleep Commands B 1 SO Sleep Options B 0 SM Sleep Mode D E460 ST Wake Period E 1770 SP Cyclic Sleep Period Change networl COMS 9600 8 N 1 FLOW NONE XBPO9 DM Figure 25 Modem Configuration Screen for X CTU Software H 1 DC IN Figure 26 Air Sensor Node Power Circuit Schematic H 2 Engineering Manual Hardware Development Purpose of Manual The engineering manual is meant to serve as a guide in the replication of a network of air quality sensors The manual is broken up into several sections each describing the details of a specific piece of hardware These sections are listed below Introduction to the System Node Dylos DC1100 Pro Laser Particle Counter XBee Pro DigiMesh 900 Module and Interface Boards Null Modem Serial Cable RS 232 Interface Voltage Regulator Circuit Power Source 9V Battery Completion of System Node Oe ee NS 1 Introduction to the System Node The completed system node is depicted in Figure 1 below and consists of a Dylos DC1100 Pro laser particle counter an XBee Pro DigiMesh 900 wireless module a voltage regulator circuit and a 9V battery supply The Dylos DC1100 sensor contains two channels for capturing particle counts The Dylos also contains an RS 232 interface allowing the data from each channel to be transmitted to the XBee Pro module The XBee Pro module can wirelessly transmit the same data to a coordinat
16. gt 50 amp amp humidityInt lt 59 conc25 conc25First humidity 100 rain50 concl0 cone25 concLarge goto stop Humidity between 60 69 else if humidityInt gt 60 amp amp humidityInt lt 69 conc25 conc25First humidity 100 rain60 concl0 cone25 concLarge goto stop Humidity between 70 79 else if humidityInt gt 70 amp amp humidityInt lt 79 conc25 conc25First humidity 100 rain70 concl0 conce25 concLarge goto stop Humidity between 80 89 else if humidityInt gt 80 amp amp humidityInt lt 89 conc25 conc25First humidity 100 rain80 concl0 conce25 concLarge goto stop Humidity between 90 100 else if humidityInt gt 90 amp amp humidityInt lt 100 conc25 conc25First humidity 100 rain90 concl0 cone25 concLarge goto stop code S1 Rain lt lt endl else cout lt lt Error see humidity correction section of else if condition NO Rain Table Improvement Factor double factor0 13 0 39 double factor40 8 40 49 double factor50 6 50 59 double factor60 4 60 69 double factor70 1 75 70 79 double factor80 1 5 80 89 double factor90 1 90 100 Humidity between 0 39 if humidityInt gt 0 amp amp humidityInt lt 39 conc25 conc25First humidity 100 factor0 concl0 cone25 concLarge goto stop Humi
17. rather than purchased from a vendor Despite the emphasis on producing a low cost system the design team does not feel it has sacrificed quality in any component of the system Manufacturability Because the system was intended to be deployed outdoors the final fabrication was constantly in mind during the design process From the earliest stages the team understood that a small to mid size weatherproof enclosure would provide the final housing for the system thus this constraint was accounted for in the evaluation and selection of system components Large or unwieldy components that were not conducive to the final fabrication goals were not considered Furthermore the design was kept simple and compact to ensure the success of the final fabrication and future reproductions Sustainability The system was designed for a long lifespan and contains no disposable parts this design goal played a large role in the decision to use a rechargeable battery as the system s main power source Furthermore the voltage regulator circuit reduces wear on the system components for instance the Dylos sensor s fan unit ensuring that they are powered on for only a fraction of each hour Environmental As previously noted the environmental impacts on the system were accounted for in the design process At the same time environmental constraints also played a role in the physical deployment of the system to avoid exposure to rain it was determ
18. EEEE EEEE ICC GIS CCGG IKI IIE CI Name Data Extraction Tool and Algorithm Application Description Overall this tool will be used to extract the useful information from the RealTerm output text file and apply this data to a particle count to mass algorithm The output of this program will be saved in a new text file This file will contain the timestamp unique C 3 node identifier node location small large particle counts count 01ft 3 and concentration ug m 3 Creator Anthony McClellan Date Created 3 31 09 Data Modified 5 12 09 Additional Notes for User 1 Need to update Node Location Table in Part 4 with actual node locations PRB RBBB ASE OSE EEEE EEEE EEEE GI ICICI ICI ICICI ICI ICI ICI ICI II I CI CS A a I CI a a CAE aE a I a include lt iostream gt include lt iomanip gt include lt fstream gt include lt string gt include lt math h gt include lt stdio h gt include lt stdlib h gt include lt ctime gt include lt time h gt include lt direct h gt include lt windows h gt include lt cstdlib gt using namespace std main program int main Information for User cout lt lt Data Extraction Tool and Algorithm Application lt lt endl lt lt endl cout lt lt Description of Program lt lt endl cout lt lt This tool will read in the RealTerm text file and extact the lt lt end cout lt lt usef
19. Select sheet Sheets Location 3 Data Select Clear current data in sheet Sheets Location 3 Data Cells Clear Begin copying file on first cell Range A1 Select Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCell Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results Location 3 Data txt Open FName For Input Access Read As 4 Sep w Use while loop to copy text file into worksheet While Not EOF 4 Line Input 4 WholeLine If Right WholeLine 1 lt gt Sep Then WholeLine WholeLine amp Sep End If ColNdx SaveColNdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value TempVal Pos NextPos 1 ColNdx ColNdx 1 NextPos InStr Pos WholeLine Sep C 30 Wend RowNdx RowNdx 1 Wend Close 4 Format Time Columns C C Select Selection NumberFormat 409 m d yy h mm AM PM Go to first cell Range A1 Select USK 2 2 CC CE CIC IC 2 KK EEEE EEE 2 KK I a ee aK oi aK aK ak Part le Import Location 4 Data txt to sheet Location 4 Data EEEE 21 IC IC IE IC a CIC C1 A KE 1 IC IC 12 I CI A A aK IC I a IC aI a AR Ca ak Ca a a Select sheet Sheets Location 4 Data Select Clear current data in sheet Sheets Location 4 Data Cells Clear Begin copying file on first c
20. Sep mn Use while loop to copy text file into worksheet While Not EOF 7 Line Input 7 WholeLine If Right WholeLine 1 lt gt Sep Then WholeLine WholeLine amp Sep End If CoINdx SaveColINdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value TempVal Pos NextPos 1 CoINdx CoINdx 1 NextPos InStr Pos WholeLine Sep Wend RowNdx RowNdx 1 Wend Close 7 Format Time Columns B B Select Selection NumberFormat 409 m d yy h mm AM PM Go to first cell Range A1 Select End of PART 1 13K 2 CC IC IC KCC ICC ICE 2 IC SIC EE KI IC I 2 aK I I a ee ok ak 2 ak aK a USK KR 2 I ICC IC C2 ICC IC KK ICC EK aK KC 2 a eK aK eo ak a 2 ok ok PART 2 Print Warning Sheet This sheet will contain any reading when PM2 5 gt 35 and PM10 gt 150 USK 2 2 EE 2 CC IC CCC IC AC IE KK IC EREEREER KC AC a eK aK eo aK a i ok ok Select sheet Sheets Warning Report Select Clear current data in sheet Sheets Warning Report Cells Clear C 33 Begin copying file on first cell Range A1 Select Copy All Data txt and filter message for lt gt Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCell Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results Warning Report txt Open
21. Tools Help Gr X X i J gt Search gt Folders A Folder Sync Address la C Documents and Settings Anthony Desktop C Results File and Folder Tasks Make a new folder Publish this folder to the Web Share this folder Other Places Desktop My Documents Shared Documents My Computer My Network Details C Results File Folder Date Modified Sunday May 10 2009 7 33 PM R ohiects Index WONAuUbWNE File Edit Format View Help All Data WDay Date Time s Tue Tue TUe Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue Tue 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 13 4 28 2009 14 4 28 2009 14 4 28 2009 14 4 28 2009 14 4 28 2009 14 4 28 2009 14 4 28 2009 14 4 28 2009 14 4 28 2009 14 All Data Text Document 337 KB Location 1 Data Text Document 121 KB Location 3 Data Text Document 117 KB Location 5 Data Text Document 1KB my my y y Averages Text Document 2 KB Location 2 Data Text Document 101 KB Location 4 Data Text Document 1 KB Warning Report Text Document 22 KB Node Location Garden 8th and mercy St HCC Roof 2 Facing Snyder Garden 8th and mercy St HCC Roof 2 F
22. as ty vk F LAA i 4 28 09 12 00 AM 4 29 09 12 00 AM 4 30 09 12 00 AM 5 1 09 12 00 AM 5 2 09 12 00 AM 5 3 09 12 00 AM 5 4 09 12 00 AM Figure 6 Continuous Data From Southeast Philadelphia System Deployment Location WDay Year Month Day PM 2 5 Average Table 4 Southeast Philadelphia Daily Particulate Matter Averages While the EPA daily average condition is met for each day of the deployment period Figure 6 shows the usefulness of the preliminary screening system as a continuous monitor For example during a short time period on May 1 2009 concentration levels become elevated to within 5 ug m of the EPA threshold Because the EPA only specifies and using filters only tests for a daily average particulate matter concentration spikes may not be noticed Using the preliminary screening system as a continuous monitor allows users to pinpoint exactly when and where short term concentration spikes may have occurred For example after receiving several weeks of data a user may notice repeated concentration spikes the public s repeated exposure to these elevated particulate matter levels could potentially be a cause for concern and should be investigated by the appropriate authority At the same time if only a few spikes are noticed the events triggering these elevated levels could be evaluated as one time events fire building demolition etc that are not a cause for concern The high particulate matter concentration sp
23. int Loc2Index 1 int Loc3 Index 1 int Loc4Index 1 int LocSIndex 1 C 8 Nariables to save previous values Location 1 int countLoc Data 1 int countLoc1Cycle 1 double sumSmallLocl 0 double sumLargeLoc1 0 string prevYear1 string prevMonth1 string prevDay 1 Location 2 int countLoc2Data 1 int countLoc2Cycle 1 double sumSmallLoc2 0 double sumLargeLoc2 0 string prevYear2 string prevMonth 2 string prevDay 2 Location 3 int countLoc3 Data 1 int countLoc3Cycle 1 double sumSmallLoc3 0 double sumLargeLoc3 0 string prevYear3 string prevMonth3 string prevDay3 Location 4 int countLoc4Data 1 int countLoc4Cycle 1 double sumSmallLoc4 0 double sumLargeLoc4 0 string prevYear4 string prevMonth4 string prevDay4 Location 5 int countLoc5 Data 1 int countLocS5Cycle 1 double sumSmallLoc5 0 double sumLargeLoc5 0 string prevYearS string prevMonth5 string prevDayS string condition PRR E ISSR RCC CICE EEE EEE EEE EEE EE EEEE EE E SCISSOR ICE Part 2 Exclude useless lines from RealTerm Output These lines show only the sensor label Dylos DC1100 and NO data EEEE EEE EEEE EEE EEEE EEEE EEEE EEEE EE E IS CIS CISC ICICI CIR IIE Read text file until end of file while input eof C 9 string line getline input line Exclude lines ending in Dylos DC1100 int dylosPos line find Dylos if dylosPos
24. lt left lt lt Conditions lt lt setw 13 lt lt left lt lt Small Count lt lt setw 13 lt lt left lt lt Large Count lt lt setw 14 lt lt left lt lt Conc PM 2 5 lt lt setw 13 lt lt left lt lt Conc PM 10 lt lt setw 8 lt lt left lt lt Message lt lt endl Location 2 ofstream outputLoc2 outputLoc2 open C Documents and Settings Anthony Desktop C Results Location 2 Data txt outputLoc2 Title lt lt setw 16 lt lt left lt lt Location 2 Data lt lt endl lt lt endl Header Header lt lt setw 6 lt lt left lt lt Index lt lt setw 6 lt lt left lt lt WDay lt lt setw 17 lt lt left lt lt Date Time lt lt setw 6 lt lt left lt lt Node lt lt setw 36 lt lt left lt lt Node Location lt lt setw 13 lt lt left lt lt Humidity lt lt setw 12 lt lt left lt lt Conditions lt lt setw 13 lt lt left lt lt Small Count lt lt setw 13 lt lt left lt lt Large Count lt lt setw 14 lt lt left lt lt Conc PM 2 5 lt lt setw 13 lt lt left lt lt Conc PM 10 lt lt setw 8 lt lt left lt lt Message lt lt endl Location 3 ofstream outputLoc3 outputLoc3 open C Documents and Settings Anthony Desktop C Results Location 3 Data txt outputLoc3 Title lt lt setw 16 lt lt left lt lt Location 3 Data lt lt endl lt l
25. of the design team s sensors Although the amount of filter data collected was limited to 12 daily readings it was still vital to compare the design team s continuous data as well as the AMS continuous data to these Federal Regulatory Method readings Figure 17 and Table 5 Appendix D show the comparison of the daily average of the continuous monitors to the daily filter readings based on this small sample the design team s system outperformed the AMS continuous monitor Though the deployment at the AMS facility was crucial in the finalization of the design team s system there were several key issues that must be mentioned e Only one month of data during one season was collected due to the time restrictions of the design project The AMS has collected data for several years during all seasons yet is still working to improve its own continuous monitoring methods e The AMS did not allow the design team access to its PM 9 monitors As a result further work can be done to collect and evaluate PM 0 data e The design team sought to minimize false alarms or instances where its data exceeded the EPA limit of 35 g m but the AMS monitors did not During their respective deployment periods Sensor 1 operated at a false alarm rate of 0 76 113 false alarms out of 14 933 total readings while Sensor 2 operated at a false alarm rate of 23 17 false alarms out of 7 255 total readings e The design team s sensor package proved to
26. the Graphical Analysis program located on the desktop The user must then run the program by clicking the macro enabled button Import All Data This is shown in Figure 23 Results PM Data and Graphs This sheet contains hyperlinks to quickly search through all documents Location Table Location 1 Not Assigned cl k t Location 2 Not Assigned Ic o Locations Not Assigned y O Import All Data Location 4 Not Assigned Location 5 Not Assigned Daily Averages Link Table Click Link to jump to sheet View Warning Report All Data Data PM 2 5 Graph PM 10 Graph Contains all readings that exceed the following thresholds Location 1 Data PM 2 5 Graph PM 10 Graph If PM 2 5 greater than 35ug m 3 Location 2 Data PM 2 5 Graph PM 10 Graph If PM 10 is greater than 150ug m 3 Data PM 2 5 Graph PM 10 Graph Location 4 Data PM 2 5 Graph PM 10 Graph Location 5 Data PM 2 5 Graph PM 10 Graph l Figure 23 Graphical Analysis Program User Controls Output Files The outputs of this program are numerous Excel worksheets within one workbook The first sheet contains the location names as well as hyperlinks for easy access to other sheets This is shown in Figure 24 These other sheets include a warning report Figure 25 an average report Figure 26 and also an array of graphs for each location Figure 27 Results PM Data and Graphs This sheet contains hyperlinks to quickly search through all
27. the purchase of mounting devices as zip ties were used to hang system components The team did purchase a second Asus Eee PC which was deployed at the AMS Site Teamwork The design team worked very well together over the course of the Senior Design Project in both group and individual sessions The team members interacted and learned from not only one another but also two outstanding faculty advisors a knowledgeable and helpful graduate advisor representatives from the Clean Air Council Environmental Protection Agency and product vendors and a team of motivated high school students from the Science Leadership Academy Anthony McClellan served as the team s leader overseeing all aspects of the design process and ensuring that all responsibilities were delegated and internal deliverables met Additionally he worked tirelessly to develop and perfect the extensive software package needed to collect and analyze the massive amount of data collected Mark Uva served as the team s wireless component expert he evaluated a variety of networking options before determining the perfect solution for the task at hand He also oversaw the budget and ensured that savings were locked in where possible enabling the final design to be produced for less than the initial estimate Edward Ostapowicz oversaw the system s power specifications including the power budget power source evaluation and current draw analysis Furthermore he designed and fabr
28. the team needed to be sure that it did not overstep any bounds in its design development and deployment For example the design team needed to get approval by the Air Management Service to observe its data and current testing equipment and could only operate onsite under their supervision and specification Most importantly in collecting data the design team needed to drive home the point that the preliminary screening system was developed to protect the public and supplement current testing procedures rather than compete with or show up the current government testing 10 Timeline The progression of the design and testing process is shown in Appendix A The majority of spring term work was spent deploying the system at the AMS Site and in Southeast Philadelphia analyzing the data collected finalizing the conversion algorithm and writing the software code needed to capture and present the data Additionally considerable time was spent working with high school students from the Science Leadership Academy This time took the form of weekly meeting in which the design team taught the students about the engineering process and helped the students develop their own system of testing for deployment in their school Budget The system design budget and industrial budget are captured in Appendix B The system was designed and fabricated for 8 less than the initial estimate of 4 000 a savings of 335 The final system did not require
29. 125 18 Small particle Count Large Particle Count File Edit Format View Help 1240923659 12 2 k ypcDylos DC1100 A 1240923659 U 22 Od GU GypcDylos DC1100 1240923659 k oy Z Dylos DC1100 1240923668 2 k ypr 546 31 1240923718 5 u fyprso0 49 1240923719 n u dy Zpylos 0c1100 1240924259 k ypcDylos DC1100 1240924259 U kK G6ypcDylos DC1100 1240924267 k yp7t588 39 1240924318 E u qypr798 48 1240924318 _ Gk oy Zpylos pc1100 1240924858 G u GypcDylos DC1100 1240924858 k yplcDylos Dc1100 1240924859 U k 6ypr874 55 1240924917 E u dy 2823 45 1240924917 m K p1570 42 1240924917 gt u dy ZDylos DC1100 1240925458 kK 6ypcDylos DC1100 1240925459 I Gk y Dylos DC1100 1240925459 U u qyprso9 50 1240925517 m DE a x y 2556 38 1240925517 oo a oypl841 61 Figure 17 RealTerm Output File result txt PHASE 2 Data Conversion and Analysis J 16 IMPORTANT Before proceeding please verify that the RealTerm output file results txt is located in the folder RealTerm Output which should be located on the desktop Programs Folders i p Import RealTerm Weather Data Output E Weather Data Output Extraction and E Algorithm C Results A Graphical Excel Output Analysis Figure 18 All Phase 2 Programs and Output Folders Located on the Desktop 1 Import Weather Da
30. 2 2A A 2A 2A I 9 C2 EE 2A A 2 2A KC 2 9 2 A ak 22 A a Ka 2 2 ak 2 ak aK a aK ak ok EndMacro On Error GoTo 0 Application ScreenUpdating True Close 8 TUTTI eee END ImportTextFile Macro TUTTI eee End Sub C 35 Appendix D AMS Site Deployment Comparison of AMS Data to ECE 19 Uncorrected Data 60 gt AMS Continuous PM2 5 MetOne H AMS Filter PM2 5 24Hr Average 50 ECE Team 19 Original Algorithm Sensor1 PM2 5 4 5 09 0 03 4 10 09 0 03 4 15 09 0 03 4 20 09 0 03 4 25 09 0 03 4 30 09 0 03 5 5 09 0 03 Figure 13 Comparison of AMS Data to ECE 19 Uncorrected Data LA afi mM E W M 7 BEYI if ah UAE a U W 4 0 r r r r r r 4 3 2009 0 54 4 8 2009 0 54 4 13 2009 0 54 4 18 2009 0 54 4 23 2009 0 54 4 28 2009 0 54 5 3 2009 0 54 120 Figure 14 Humidity Data During Deployment Comparison of AMS Data to ECE 19 Corrected Data 60 gt AMS Continuous PMZ 5 MetOne AMS Filter PM2 5 24Hr Average ECE Team 19 Sensor1 PM2 5 uw o rS o 1 i W Concentration ug m N w Oo Oo o fe ea pa dies deal hia 4 5 09 0 03 4 10 09 0 03 4 15 09 0 03 4 20 09 0 03 4 25 09 0 03 4 30 09 0 03 5 5 09 0 03 Figure 15 Comparison of AMS Data to ECE 19 Corrected Data Comparison of AMS Data to ECE 19 Corrected Data Sensor 2 gt a AMS Continuous PM2 5 MetOne ECE Team 19 Sensor 2
31. 2009 Fri 2009 Sat 2009 Sat 2009 Sat 2009 Sun 2009 Sun 2009 Sun 2009 29 29 29 30 30 aOnuunnunununun fk A A A A A Ww oO WWwWwWwnNnNNE BB 4 61978 5 95159 5 37765 4 15855 3 47878 3 73218 7 12313 5 81711 6 55181 19 1732 18 4497 19 2174 8 59338 8 43104 7 67917 PM 10 Average Message 19 2471 24 5803 22 5017 15 892 16 6734 15 5493 50 4017 51 772 53 7101 143 84 117 262 108 316 22 7369 37 4068 23 1048 Figure 26 Graphical Analysis Output Average Report J 22 4 28 09 12 00 AM 4 29 09 12 00 AM 4 30 09 12 00 AM 5 1 09 12 00 AM 5 2 09 12 00 AM 5 3 09 12 00 AM 5 4 09 12 00 AM 5 5 09 12 00 AM t Location 1 PM2 5 t Location 2 PM 2 5 Location 3 PM 2 5 Location 4 PM 2 5 Location 5 PM 2 5 Figure 27 Graphical Analysis Output All Locations PM2 5 J 23 Anthony McClellan 15 Elmwood Avenue Norwood PA 19074 C 484 574 2164 anthony s mcclellan drexel edu anthony meclellan gmail com Education Drexel University Philadelphia PA Graduation June 2009 Bachelor of Science in Electrical Engineering Minor in Business Administration Cumulative GPA 3 4 Skills e Software Microsoft Office Microsoft Visio AutoCad MatLab PSpice LabView Maple Zuken E3 Cable Tool e Programming Languages Basics of HTML C Java and VDHL Engineering Design Projects Improved various skills including leadership organization and research e Relev
32. 4041 12 7526 13 0985 15 3813 13 079 13 3369 17 6005 19 7933 20 7155 20 9158 17 197 19 8345 20 0541 16 4835 17 8859 16 983 164 553 Warning PM 10 too high 181 814 Warning PM 10 too high 181 217 Warning PM 10 too high 164 6 Warning PM 10 too high 150 395 Warning PM 10 too high 199 794 Warning PM 10 too high 169 252 Warning PM 10 too high 161 668 Warning PM 10 too high 203 552 Warning PM 10 too high 166 059 Warning PM 10 too high 191 343 Warning PM 10 too high 195 252 Warning PM 10 too high 159 71 Warning PM 10 too high 178 786 Warning PM 10 too high 199 091 Warning PM 10 too high 196 151 Warning PM 10 too high 170 202 Warning PM 10 too high 201 809 Warning PM 10 too high 192 916 Warning PM 10 too high 206 169 Warning PM 10 too high Figure 25 Graphical Analysis Output Warning Report Location Garden 8th and Mercy St HCC Roof 1 Facing 8th St HCC Roof 1 Facing 8th St Garden 8th and Mercy St HCC Roof 1 Facing 8th St Garden 8th and Mercy St Garden 8th and Mercy St HCC Roof 1 Facing 8th St HCC Roof 1 Facing 8th St Garden 8th and Mercy St Daily Averages all Locations l HCC Roof 2 Facing Snyder St HCC Roof 2 Facing Snyder St HCC Roof 2 Facing Snyder St HCC Roof 2 Facing Snyder St HCC Roof 2 Facing Snyder St WDay Year Month Day PM 2 5 Average Wed 2009 Wed 2009 Wed 2009 Thu 2009 Thu 2009 Thu 2009 Fri 2009 Fri
33. 5 e Researched several technologies considered for use in securing the United States borders Delegated tasks integrated separate design components and assembled final document for submission Presented design before a panel of faculty and students e Selected for presentation at the 85th Annual Meeting of the Transportation Research Board in Washington DC e Published in the 2006 edition of the Transportation Research Record High Profile Public Speaking Experience Democratic Presidential Debate Drexel University Co Host October 2007 e Emceed viewing party attended by students and world renowned political figures e Greeted and introduced Democratic Presidential Candidates on stage Anthony J Drexel Society Gala Drexel University Student Ambassador November 2007 e Selected to participate in Drexel University s annual gala event e Dined and interacted with prominent university benefactors and donors Discover Drexel Day Open House Drexel University Emcee and Keynote Speaker October 2007 Delivered welcoming speech to over 500 high ability visiting students Introduced guest speakers and provided direction for execution of tasks Software Skills and Proficiencies e Microsoft Office MATLAB AutoCAD e Maple 9 5 e Lab View e PSPICE Mark Uva 3175 JFK Blvd Apt 814 Philadelphia PA 19104 973 727 3321 mark a uva drexel edu Drexel University Philadelphia PA Bachelor of Science in Electrical Engineering Anticipated Gra
34. 5 lt lt lt lt setw 20 lt lt left lt lt messageLoc5 lt lt endl countLocSCycle countLoc5Data 0 sumSmallLoc5 0 sumLargeLoc5 0 Average Calculations countLoc5Data Count total entries for average sumSmallLoc5 cone25 Sum Concentration of 2 5 sumLargeLoc5 conc10 Sum Concentration of 10 Apply previous variables for average calculation prevYear5 yearID prevMonthS monthID prevDay5S dayID End If End If End If End If End While Close files C 26 input close outputAllData close outputLocl close outputLoc2 close outputLoc3 close outputLoc4 close outputLoc5 close averages close Program Complete cout lt lt Program Complete lt lt endl cout lt lt Now execute Graphical Analysis Program lt lt endl lt lt endl return 0 Graphical Analysis Code Type VBA Macro Program Name PM Data Updater Description This macro will need to be run once all PM text files are updated These files are kept in the folder labeled C Results Creator Anthony McClellan Date of last update 5 12 09 ImportTextFile referenced from http www cpearson com excel ImpText aspx Sub Graphical _Analysis U3 2 2 CE KC CEE CC CCE I IC CE I IC Ee 2 E I eo aK eo ak 2 a 2 ok ok PART 1 Import all text files i This data is located in the c results folder on the desktop EEEE 4 1 9 E22 EE 2A 2A I 9 2
35. A A E22 9 E22 E E E I 9 C2 EEEE EEE C2 24 2 A a Ka 2 ak a 2 a ok V2 2 2K 9 4 1 9 E22 9 22 A I 9 ERER EC 2 9 2 2A a 2A 91 9 Ko 2 2 2 ak ak ok ak ak a ak Part la Import All Data txt to sheet All Data Vf 2 1 4 9 E22 2 2A EEEE ER E 2 9 C2 2A 2A A 9 ok 2 2 2 22 2 ak ak ok ak ak a ake Update User name for each computer Dim UserName As String UserName Anthony Select sheet Sheets All Data Select Clear current data in sheet Sheets All Data Cells Clear Begin copying file on first cell Range A1 Select All variables Dim RowNdx As Long Dim ColNdx As Integer Dim TempVal As Variant Dim WholeLine As String Dim Pos As Integer Dim NextPos As Integer Dim SaveColNdx As Integer C 27 Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCell Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results All Data txt Open FName For Input Access Read As 1 Sep mn Use while loop to copy text file into worksheet While Not EOF 1 Line Input 1 WholeLine If Right WholeLine 1 lt gt Sep Then WholeLine WholeLine amp Sep End If CoINdx SaveColINdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value Temp Val Pos NextPos 1 CoINdx CoINdx 1 NextPos InStr Pos WholeLine Sep Wend RowNdx R
36. E EEE EEEE EE EE EEEE EEEE IOC GS ICSC ICICI IIRC Read in Weather File string ext csv string dash string weather Weather string directory C Documents and Settings Anthony Desktop Weather Output Output from Import Weather Data string fileName directory weather yearID dash monthID dash dayID ext double humidity ifstream weatherFile weatherFile open fileName c_str ios binary if weatherFile fail cout lt lt Weather error lt lt endl cout lt lt Weather file does not exist Please FIRST run Import Weather Data for all dates needed lt lt endl cout lt lt Program will now end lt lt endl C 14 return 0 Read text file until end of file while weatherFile eof string weatherLine getline weatherFile weatherLine Exclude header line int headerLoc weatherLine find Time if headerLoc 1 Find weather hour int hourFind weatherLine find string wHour weatherLine substr 0 hourFind convert hour to int int hourFinalInt atoi wHour c_str Ideal Case IF weather hour matches hour for sensor Alternative Take humidity reading from next hour if hourF inalInt hourIDInt hourFinalInt 1 hourIDInt if hourFinalInt hourIDInt Comma Locations 1 4 Comma 1 Location int comma Loc weatherLine find Comma 2 Location int comma2Loc weatherLine find
37. E tot sas E E aad ach Saisauai a dyer uacaus merece Salone C 27 Appendix D AMS Site Deployment 0 2cc c secs sceansteccsscaveteovhadsesesssauseaeittesscrveteeens D 1 Appendix E Southeast Philadelphia Deployment s snssssessesessseossessseseossessrsseesseesees E 1 Appendix F EPICS Mentoring seessessssssesseessossrsseossessrsseessossesseessessosseesseesossesseonoeso F 1 Appendix G System Hardware Technical Discussion cccccesccesseeeseeesseeeeeeteeeeees G 1 ofA S101 se a DD OTS SEMEN SOE SOE A a E a SRSLY E E E a G 1 INC WOT Arise see uea a oe all ri Geeta E E r A aaa G 2 POW CD sis de eet ise Sid eases ad ee A ae Ne G 3 Node Placement apace ccs sts esir netni anenai riean aaa ta Eie a Aniara ANR G 3 Appendix H Additional Figures sseseseseessessesessseessesssssressessrssrossessrssresseeseeseesseesees H 1 Appendix I Engineering Miarial detivs laos sess saadnds ca pnis sacsonk eneseeataischenvayacsines I 1 Appendix J User Wana iy4 1 es mensuereiorsia sie eau EE EEE TTE J 1 Appendix K Resumes chsesdscnat innur ua a Arete tidnd ae asian aa ew Aedes K 1 iii List of Figures Figure 1 Overview of System Data Flow and Actual Node Configuration 1 FPisure 2 Dyvios DCT IOO PrO Gace a e a E EA AE AA ESS 3 Fig re 3 Xbee Pro DigiMesh 900 saiia a a a ee i a aaa 3 Figure 4 Base Station As s Eee PCO ron apuse tetes oi aacuses EO au SERO 3 Figure 5 Impact BDP 8000 Batters ss cws sne
38. EE EEE EEEE EEEE EEEE EEEE EE E ICIS ISIC CII IRE Part 4 Node Location Lookup Table Connects node identifier output file to physical placements of the device EEEE EEE EEEE EE EEEE EEEE EE EEE E EEEE EEEE IC GIS CISC ICICI IEICE This table will be updated once all node identifiers and all node locations have been identifier string location int locationNumber 0 Type of Deployment AMS Location or Demo if choice 1 if nodeldentifier H location Under MetOne Sensor locationNumber 1 else if nodeldentifier H location Attached to MetOne Sensor locationNumber 2 Houston Community Center Set Up if choice 2 if nodeldentifier K location Garden 8th and Mercy St locationNumber 1 else if nodeldentifier K location HCC Roof 1 Facing 8th St locationNumber 2 else if nodeldentifier U location HCC Roof 2 Facing Snyder St locationNumber 3 else if nodeldentifier C location Location 4 locationNumber 4 Uo else if nodeldentifier D location Location 5 locationNumber 5 else location ERROR PRR E ECCI R EEEE EEE EEEE EEEE EEEE EEEE EE EEE SCISSOR ICICI IRE Part 5a Conversion Algorithm Particle Count per 0 1ft 3 to Mass per m 3 Original Rob Falcone Updates Anthony McClell
39. FName For Input Access Read As 8 Sep Wn Use while loop to copy text file into worksheet While Not EOF 8 Line Input 8 WholeLine If Right WholeLine 1 lt gt Sep Then WholeLine WholeLine amp Sep End If CoINdx SaveColINdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value Temp Val Pos NextPos 1 CoINdx CoINdx 1 NextPos InStr Pos WholeLine Sep Wend RowNdx RowNdx 1 Wend Go to first cell Range A1 Select End of PART 2 13K 2 2 KC IC ICICI IC ICCC KIC ICCC IC IC EE KIC IC I 2 KC KIC a ee ok ak 2 2 aK aK a 13K 2 2 IC 2 CACC I 2 CC ICE KK ICRC IE Ka KIC RC ae aK aK eo ak a eo ok ok PART 3 Copy all sheets to new workbook 13K 2 2 EC EC I ICCC KC ICRC I eK aK KCC a eK aK eo aK a eo ok ok Dim sAppPath As String sFileName As String sDate As String sAppPath C Documents and Settings amp UserName amp Desktop Excel Output sDate Replace FormatDateTime Now vbShortDate _ sFileName sAppPath amp Excel Output amp sDate amp xlsm ActiveWorkbook SaveAs Filename _ sFileName FileFormat _ xlOpenXMLWorkbookMacroEnabled Password WriteResPassword ReadOnlyRecommended False _ CreateBackup False Select sheet C 34 Sheets User Info Select Go to first cell Range A1 Select End of PART 3 V2 2 1 4 9 E22 EE 2A 2 2A I EC 2A E E E A 2A 9A EE 2 C
40. Figure 4 a Cable Components 1 DB 9 Male Connectors e Manufacturer Tyco Electronics e Part Number 5747250 4 2 Wiring b Null Modem Cable Construction a Connect pin 5 ground from one header to pin 5 of the other header via wire wrapping or soldering b Connect pin 2 in of header 1 to Pin 3 out of header 2 via wire wrapping or soldering c Connect pin 3 out of header to pin 2 in of header 2 via wire wrapping or soldering 1 4 lO i BG ee o E EE RS 232 5 GROUND 3 DATA OUT TXD 2 DATA IN RXD Figure 4 RS 232 DB 9 Pin Out 5 Voltage Regulator Circuit The voltage regulator circuit controls the power supplied to the Dylos air quality sensor It allows power to be supplied to the sensor only when a reading is taken For all battery powered nodes this circuit must be utilized due to the high power consumption of the sensor a Circuit Components a Voltage Regulator Chip See Figure 5 i Manufacturer Sharp Microelectronics ii Part Number PQO90RDA1SZH b 33 uF Capacitor i Manufacturer Panasonic ECG ii Part Number ECA 1HHGR33 c 47 uF Capacitor i Manufacturer United Chemi Con ii Part Number EKY 250ELL470ME11D d 9 V power jack i Manufacturer CUI Inc ii Part Number PJ 018H e 9 V adapter with cable assembly i Manufacturer Tensility International Corp ii Part Number CA 2188 f PC Board i Manufacturer Vector Electronics ii Part Number 8015 1
41. Figure 4 Base Station Asus Eee PC determined and Figure 3 Xbee Pro DigiMesh 900 Figure 5 Impact BPD 8000 Battery Deliverables e Hardware o Multiple node network minimum of 4 Each node consists of an air sensor and wireless networking component o Base station to collect data e Software o Networking o Data collection and storage o Data analysis algorithm Convert data collected by air sensors into appropriate form for comparison to EPA Standard e Documentation o User Manuals Written to inform client on system use o Engineering Manuals Written to inform future engineers who may build upon system e Additional Deliverables that Have Been Added o Design and fabrication of node enclosures o Comparison of collected data to EPA AMS collected data AMS Test Deployment Calibration and Validation The design team worked with the Air Management Service AMS to deploy two sensors at the AMS Lycoming Avenue testing site here comparison to the Environmental Protection Agency s accepted methods of measurement for PM25 took place This experience allowed the team to determine its sensors accuracy as well as calibrate and improve the derived particle count to concentration conversion algorithm specific details regarding the algorithm can be found in the next section There are several important notes regarding the AMS deployment At the site the design team was able to compare its system to the EPA Federal Reference Me
42. I 5 29 1 050RDA2 SHARP x identification mark 3112 54 fa Cay CaN ia VDI Figure 5 Voltage Regulator Schematic b Construction of Voltage Regulator Circuit re ios oe Prepare PC board by modifying the size to approximately lin x 1 in Review circuit schematic in Figure 6 and before soldering components to the PC board Review Figure 7 to see an example of proper placement of components on the PC board Solder components voltage regulator chip and capacitors to PC Board A connection to XBee Pro serial interface board must be made to control power supplied to the Dylos sensor i First make a connection to pin 4 on the voltage regulator chip ii Second make a connection to the XBee Pro serial interface board to the right side of the 4 resistor See Figure 7 for an example of proper placement Attach a 9V power adapters to the board by using the attached twisted pair cable assembly i Dylos sensor 9V power connection positive end of twisted pair should be connected to pin 2 of the voltage regulator chip ii XBee Pro 9V power connection positive end of the twisted pair should be connected to pin 1 of the voltage regulator chip iii Negative ends of both twisted pairs must be grounded to a common ground on the PC Board Mount the completed circuit to the XBee Pro module by using a screw 1 6 Figure 6 Complete Voltage Regulator Circuit Schematic F
43. IR CI Add Message to each line If PM2 5 gt 35 or PM10 gt 150 a warning message is displayed If Lower then nothing is added string message Warn 3 both PM2 5 and PM10 over threshold if conc25 gt 35 amp amp conc10 gt 150 message Warning PM 2 5 and PM 10 too high Warn 1 PM2 5 over threshold elseif conc25 gt 35 message Warning PM 2 5 too high Warn 2 PM10 over threshold else if conc10 gt 150 message Warning PM 10 too high No Warning else message EEEE EEEE E EEEE EEE EEE EEE EEEE EEEE EE EEEE EEE EE EEEE EEEE EE E EE E II CI Part 6 Write all data to a new files Warning Report All data Locations 1 5 PRR A EEEE EEEE EEEE EEEE EEEE EE EEEE EEE SCISSOR ICICI Exclude invalid data from sensor See in Testing C 18 if smallCountInt gt 0 amp amp largeCountInt lt 1200000000 Creation of concatDate string concatDate monthID dayID yearID Creation of concatTime string concatTime hourID minuteID Creation of concatDateTime string concatDateTime concatDate concatTime Write all information to new file All Data outputAllData lt lt setw 6 lt lt left lt lt AllDataIndex lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 16 lt lt left lt lt concatDateTime lt lt lt lt setw 5 lt lt left lt lt nodeldentifier lt lt lt lt se
44. Loc4 gt 150 messageLoc4 Warning PM 10 too high No Warning else messageLoc4 C 24 averages lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 5 lt lt left lt lt yearID NN lt lt setw 6 lt lt left lt lt monthID ae lt lt setw 4 lt lt left lt lt dayID kanm lt lt setw 20 lt lt left lt lt smallAverageLoc4 lt lt wo lt lt setw 20 lt lt left lt lt largeAverageLoc4 lt lt lt lt setw 20 lt lt left lt lt messageLoc4 lt lt endl countLoc4Cycle countLoc4Data 0 sumSmallLoc4 0 sumLargeLoc4 0 Average Calculations countLoc4Data Count total entries for average sumSmallLoc4 cone25 Sum Concentration of 2 5 sumLargeLoc4 conc 10 Sum Concentration of 10 Apply previous variables for average calculation prevYear4 yearID prevMonth4 monthID prevDay4 daylID Location 5 else if locationNumber 5 outputLoc5 lt lt setw 6 lt lt left lt lt LocSIndex lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 16 lt lt left lt lt concatDateTime lt lt lt lt setw 5 lt lt left lt lt nodeIdentifier lt lt lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 12 lt lt left lt lt humidity lt lt lt lt setw 11 lt lt left lt lt condition lt lt
45. Met One Aerocet 212 this drawback was outweighed by the DC1100 Pro s price per unit Cost effectiveness played a large role in decision methodology since the design is a preliminary scan of ambient air 13 Device Calibration The original Dylos DC1100 Pro sensor was not calibrated to the necessary specifications As stated on the company s website The DC1100 Pro has increased lower sensitivity detecting particles down to 0 5um The large particle size range is calibrated to 2 5um and above 14 However for this specific system the Dylos Corporation agreed to calibrate the laser particle counters with a greater degree of specificity in which one channel evaluates particle sizes between 0 5um to 2 5um and the other channel evaluates particle sizes between 2 5um and 10um 4 By calibrating the device to these specifications the data can be compared to the EPA standards for PM2 5 and PMjo The disadvantage to this process is that it is very time consuming for the technician to properly calibrate each device Due to the complexity of this process the 20 price discount originally quoted in the proposed budget could no longer be granted by the Dylos Corporation Serial Interface In addition to the calibration of the device the laser particle counter was customized to include a serial data connection This simple interface allows the sensor to send out two particle counts one for each channel to the networking modu
46. Month As Integer BeginMonth Cells 13 B Value Dim BeginDay As Integer BeginDay Cells 13 C Value Dim BeginDate As String BeginDate BeginYear amp amp BeginMonth amp amp BeginDay End Date Dim EndYear As Integer EndYear Cells 18 D Value Dim EndMonth As Integer EndMonth Cells 18 B Value Dim EndDay As Integer EndDay Cells 19 C Value Dim EndDate As String EndDate EndYear amp amp EndMonth amp amp EndDay Location Information used to find the best weather based on proximity to source NE Phila or Phila Int Airports Dim Location As String Location Cells 13 G Value Dim LocationID As String If Location Philadelphia International Airport Then LocationID KPHL End If If Location NE Philadelphia Airport Then LocationID KPNE End If Save Entered Dates This could be used to enhance the algorithm file Currently not being used Sheets DO NOT CHANGE2 Select C 1 Application DisplayAlerts False ActiveSheet SaveAs Filename C Documents and Settings amp UserName amp Desktop Weather Output Entered Dates and Location Entered Dates and Location csv FileFormat xICSV _ CreateBackup False Application DisplayAlerts True Sheets User Inputs Select Loop to continue capturing weather data until the program reaches the end date Do Creation of Title used for saving the file Dim DateTitle As String DateTitle BeginYear amp
47. Navy Cruiser hulls Drexel University Philadelphia PA Executive Ambassador September 2004 Present e Promote Drexel University to prospective students and their families through guided tours e Participate in interviewing candidates for the Ambassador position All Star Baseball Academy Broomall PA Instructor May 2001 April 2007 e Instructed youth players in the fundamentals of baseball e Directed several of the Academy s instructional programs e Scheduled appointments and managed accounts for Academy clients Leadership and Organizational Experience The DAC Pack Drexel University Athletics President April 2007 Present e Increased revenue over 400 during tenure through advertising and partnership agreements Manage a budget of over 50 000 e Build a portfolio of corporate and University sponsors Draft negotiate and sign partnership contracts Design and implement creative marketing strategies and campaigns Drexel University Philadelphia PA Overnight and Day Visit Program Coordinator February 2006 March 2007 e Restructured the previous Overnight and Day Visit Program in place e Planned visits to Drexel University for selected high school seniors Fielded calls and questions from prospective students and families Trained and direct a team of Drexel University Student Hosts Engineering Design and Publication Experience The Smart Borders Engineering Design Team Drexel University Project Leader January 2005 June 200
48. PM10 over 150ug m 3 else if largeAverageLocl gt 150 messageLocl Warning PM 10 too high No Warning else messageLocl averages lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 5 lt lt left lt lt weekDayID agn lt lt setw 5 lt lt left lt lt yearID gln C 20 lt lt setw 6 lt lt left lt lt monthID lt lt lt lt setw 4 lt lt left lt lt dayID gamn lt lt setw 20 lt lt left lt lt smallAverageLoc1 lt lt wo lt lt setw 20 lt lt left lt lt largeAverageLocl lt lt lt lt setw 20 lt lt left lt lt messageLocl lt lt endl countLoc1Cyclet countLoc1 Data 0 sumSmallLoc1 0 sumLargeLocl 0 Average Calculations countLoc 1 Data Count total entries for average sumSmallLocl cone25 Sum Concentration of 2 5 sumLargeLocl conc10 Sum Concentration of 10 Apply previous variables for average calculation prevYearl yearID prevMonth1 monthID prevDayl daylID Location 2 else if locationNumber 2 outputLoc2 lt lt setw 6 lt lt left lt lt Loc2Index lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 16 lt lt left lt lt concatDateTime lt lt lt lt setw 5 lt lt left lt lt nodeIdentifier lt lt lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 12 lt lt left lt lt humidity lt lt
49. Preliminary Screening System for Ambient Air Quality in Southeast Philadelphia A Final Report submitted to Dr Timothy Kurzweg Dr Kapil Dandekar and the Senior Design Project Committee of the Electrical and Computer Engineering Department of Drexel University Team Number ECE 19 Team Members Mark Uva Electrical Engineering Robert Falcone Electrical Engineering Anthony McClellan Electrical Engineering Edward Ostapowicz Electrical Engineering Submitted in partial fulfillment of the requirements for the Senior Design Project May 13 2009 EXECUTIVE SUMMARY Detection of dangerous particulate matter in ambient air has proven to be an expensive and time consuming process As a result a low cost yet effective preliminary screening solution is necessary The system developed utilizes a laser particle counter to capture samples of ambient air Information collected by several laser particle counters is transmitted across an Xbee network to a central base station where data from all nodes each node consists of a laser particle counter and a Xbee component is collected Data is then passed through a mathematical calculation so that it can be compared directly to the National Ambient Air Quality Standard NAAQS The following milestones were reached e Fully functional system has been deployed in Southeast Philadelphia e Software to handle all aspects of data analysis and presentation has been developed o Mathematical algorithm to conve
50. Properties used to properly increase date Automatically includes leap year If BeginDay 29 And BeginMonth 2 Then BeginDay 1 BeginMonth BeginMonth 1 End If All months with 30 days If BeginDay 31 And BeginMonth 4 Or BeginDay 31 And BeginMonth 6 Or BeginDay 31 And BeginMonth 9 Or BeginDay 31 And BeginMonth 11 Then BeginDay 1 BeginMonth BeginMonth 1 End If All months with 31 days If BeginDay 32 And BeginMonth 1 Or BeginDay 32 And BeginMonth 3 Or BeginDay 32 And BeginMonth 5 Or BeginDay 32 And BeginMonth 7 Or BeginDay 32 And BeginMonth 8 Or BeginDay 32 And BeginMonth 10 Then BeginDay 1 BeginMonth BeginMonth 1 End If Increase to next year If BeginMonth 12 And BeginDay 32 Then BeginMonth 1 BeginDay 1 BeginYear BeginYear 1 End If BeginDate BeginYear amp amp BeginMonth amp amp BeginDay Loop Until BeginDay EndDay And BeginMonth EndMonth And BeginYear EndYear Automatically open algorithm program C RetVal Shell C Documents and Settings amp UserName amp Desktop Data and Analysis Data and Code Data Extraction and Algorithm Debug Data Extraction and Algorithm exe 1 Application DisplayAlerts False Application Quit End Sub Data Extraction and Algorithm Application Code Type C IMPORTANT Change all Anthony to match user name of Computer ECETeam 19 PRR E SCRE CIR ICC EEE EEEE EEE EE
51. Select USK 2 CC ICC ICCC CI IC KK ICI ICE 2 CK aK IC oa He eo ak 2 aK aK a Part 1c Import Location 2 Data txt to sheet Location 2 Data EEEE 21 1 EI C2 IE C2 1 CI CI A a IE 1 CI a IC 12 CI A A a C2 I a IC aI a AR Ca a a a a a Select sheet Sheets Location 2 Data Select Clear current data in sheet Sheets Location 2 Data Cells Clear Begin copying file on first cell Range A1 Select Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCell Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results Location 2 Data txt Open FName For Input Access Read As 3 Sep EN Use while loop to copy text file into worksheet While Not EOF 3 Line Input 3 WholeLine If Right WholeLine 1 lt gt Sep Then C 29 WholeLine WholeLine amp Sep End If CoINdx SaveColINdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value TempVal Pos NextPos 1 CoINdx CoINdx 1 NextPos InStr Pos WholeLine Sep Wend RowNdx RowNdx 1 Wend Close 3 Go to first cell Range A1 Select 13K 2 2 CE IC EE ICCC CE 2 CK IC IC EK aK a 2 2 a ee aK 2 aK aK a Part 1d Import Location 3 Data txt to sheet Location 3 Data EEEE 21 EI C2 IE IC I CIC C1 A 1 IE IC a a E E CI a A A a1 E aI a AR Ca ak Ca a ak a
52. abilities and reduce improper placement of equipment In order to utilize this useful tool a NULL is required for serial port of the interface board This NULL is provided in the XBee Pro development kit This is shown in Figure 10 Perform Test 1 2 3 4 5 6 7 8 9 Connect coordinator to USB port Open the X CTU and select the associated COM Port If the Enable APT button is checked uncheck it Verify the settings are as follows o Baud Rate 9600 Flow Control None Data Bits 8 Parity None Stop Bits 1 Make sure the modem configurations are set to the default parameters If this is not done already click the Restore button on the modem configuration page Connect a NULL to the serial port of the serial interface board as shown in Figure 10 Select the Range Test Tab Check the RSSI Checkbox to enable Received Signal Strength Indicator As shown in Figure 11 O O OO 10 Click Start to begin the test 11 Move the module away from the base station to find the maximum range 12 Click Stop to end the range test 13 The signal strength indicator is shown in Figure 12 14 Alternative to software view the LEDs on the serial interface board o One glowing green if receiving data from the coordinator Figure 10 Range Test Setup FC Sottings Range Test Teirinal Modem Configuration Cleer Stats Advanced gt gt gt Teg Loop Back Enable to display received signal stren
53. acing Snyder HCC Roof 1 Facing 8th st sHCC Roof 2 Facing Snyder Garden 8th and mercy St HCC Roof 2 Facing Snyder Garden 8th and mercy St HCC Roof 1 Facing 8th st HCC Roof 1 Facing 8th st HCC Roof 2 Facing Snyder Garden 8th and mercy St Garden 8th and Mercy St HCC Roof 2 Facing Snyder HCC Roof 1 Facing 8th st HCC Roof 2 Facing snyder Garden 8th and mercy St HCC Roof 1 Facing 8th St Garden 8th and mercy St HCC Roof 2 Facing Snyder HCC Roof 1 Facing 8th st HCC Roof 2 Facing snyder Garden 8th and mercy St st St St St St St St St Humidity Conditions Small Count Large Count 697 KR Figure 21 C Results Folder f 800 588 798 874 823 570 809 556 841 803 808 598 537 838 916 874 622 1017 684 1045 861 854 561 R Mv Comnuter NN OUOUINNRUAANEHDEANADEDS Cconc PM 2 5 57496 70324 92688 68648 32328 89595 77606 77865 65875 04678 72837 77027 01067 49955 02164 6752 32328 21177 52149 73127 7561 21436 1557 70064 Figure 22 All Data Output File J 20 3 Graphical Analysis The graphical analysis program is the final program to execute This program will automatically update an excel template file with the C Results files shown in Figure 21 User Input The user must first open
54. an PRR EEC EEEE EEE EEEE EEEE EEEE EEEE EE EEEE I E EEEE E E E C GCI CII IGE double r25 0 44 pow 10 0 6 0 um reference Lee paper double r10 2 6 pow 10 0 6 0 um reference Lee paper const double PI 3 14159 double vol25 4 0 3 0 PI pow r25 3 0 double vol10 4 0 3 0 PI pow r10 3 0 double density 1 65 pow 10 0 12 0 ug m3 reference titarelli paper double mass25 density vol25 ug double mass10 density vol10 ug dylos output particles 01 ft 3 35 315 ft 3 1 m 3 35 315 01 3531 5 m 43 double K 3531 5 per m 3 matrix 1620 58 1700 52 9000 107 1840 75 1968 85 1654 75 1730 52 PC25 matrix 1 PC10 matrix 2 PM2 5 concentration particle count from channel 1 double conc25First smallCountInt K mass25 ug m 3 double conc25 PM10 concentration particle count sum of small and large concentrations double concLarge largeCountInt K mass10 ug m 3 double conc10 PRR E IERIE EEEE EEE EE EEE EEEE EEE EE EEEE EEEE IOC GIS ICIS CICS CIR CI Part 5b Weather lookup for PM2 5 Correction Factor NOTE This section is only to improve the quantity for PM2 5 This process can not be used for PM10 since we were unable to obtain PM10 measurements from AMS IMPORTANT This section will only work properly if all weather data has been imported using the Excel Macro Import Weather Data RRA SCRE EEEE EEE E
55. ant Coursework Communication Systems Lightwave Engineering Wireless and Optical Electronics Wireless Communication Systems Modulation and Coding Digital Signal Processing numerous Design Labs Experience L 3 Communications Camden NJ Hardware Engineering Co op April 2008 Present e Received clearance in order to work on government contracts e Assisted various engineers EE ME Systems in the design of L 3 s Integrated Communication System e Gained experience with numerous types of equipment designed by both L 3 and outside vendors Internal Comms voice terminals broadcast and alarm systems External Comms Radios HF UHF VHF Networks Cameras Computer Workstations Entertainment and Training Systems Experienced broad view of L 3 s design processes High level system view of all equipment connections Lower level view of the equipment distribution by room location Lowest level view of the cable design e Utilized E3 Cabling Tool to designate appropriate mating connectors pins signal types functions e Participated in CFR Code of Federal Regulations testing procedures on an assembled communication s rack e Created various organizational tools including an obsolete parts listing and a voice switch planning tool Comcast Philadelphia PA Voice Engineer Comcast Digital Voice April 2007 to September 2007 e Assisted Market Design Engineers with design upgrading processes and equipment purchasing e Gained experien
56. ative Met One Instruments Private Communication 2009 10 Joseph Otis Minott Clean Air Council 2008 11 Dorsey R et al Simulation and Quantification of Particulate Matter Emissions in Philadelphia ECE 17 Drexel University Electrical and Computer Engineering Senior Design 2008 12 Tittarelli T et al Estimation of particle mass concentration in ambient air using a particle counter Atmospheric Environment vol 42 pp 8543 8548 2008 13 American Allergy Supply Dylos DC1100 Air Quality Monitor June 2008 http www americanallergysupply com dylos htm 14 Dylos Corporation DC1100 Pro Air Quality Monitor http dylosproducts com ornodcproair html 15 Digi International XBee Pro DigiMesh 900 Product Data Sheet 2009 http www digi com products wireless zigbee mesh xbee digimesh 900 jsp 13 16 Digi International XBee Pro DigiMesh 900 Product User Manual 2009 http ftp 1 digi com support documentation 90000903_B pdf 17 XBee Technician Digi com Private Communication 2009 18 Primex s Indoor Outdoor Enclosures 2009 http www primex ca enclosures htm 19 L Com Global Connectivity Weatherproof enclosures 2009 http www 1l com com content NEMAEnclosures html 20 Lee J et al Seasonal variations of particle size distributions of PAHs at Seoul South Korea Air Quality Atmospheric Health vol 1 pp 57 68 2008 21 Quok e
57. be much more compact than the AMS continuous monitor Figure 18 Appendix D and far less complex than the AMS filter method Figure 19 Appendix D Particle Count to Concentration Conversion Algorithm Deriving Algorithm and Code Development Because the sensor used to measure air quality is a laser particle counter the output data must be altered so that it can be directly compared to the EPA standard Table 1 Specifically an algorithm is needed to convert from particles 01 ft the output of the Dylos DC1100 Pro to ug m Similar procedures have been done previously with a high level of success 12 It is important to note that this conversion is only a strong approximation because it is impossible to quantify the exact properties of each of the thousands of microscopic particles being counted several assumptions are made in the calculation The algorithm developed assumes e All particles are spherical with a density of 1 65E12 wg m 12 e The radius of a particle in the PM2 channel is 44 um 20 e The radius of a particle in the PMjo channel is 2 60 um 20 From the above assumptions the volume and thus the mass of a particle in both size channels can be calculated approximated Multiplying the number of particles per volume by the mass per particle yields a concentration which can be converted into appropriate units ug m for comparison with the EPA standard As described in the previous section correction facto
58. cally the weather data is used to apply correction factors to the original algorithm This program will output multiple files including a warning report an average report and data reports for each location User Input The user must first run the executable file Data Extraction and Algorithm located on the desktop The user will then be asked to enter the deployment location AMS or Houston Community Center Additional deployment locations can be added but this would require updating the node locations within the source code Hix Data Extraction Tool and Algorithm Application Description of Program This tool will read in the RealTerm text file and extact the useful data This data will then be passed through a mass conversion algorithm All data will then be saved to a new file Location of RealTerm text file is in folder RealTerm Output Output format of new files in folder C Results User Directions Enter 1 AMS Enter 2 Houston Community Center User enter number from choices above Figure 21 User Input for Data Extraction and Algorithm App Output Files The outputs of this program are numerous text txt files These files are located in the C Results folder which is located on the desktop The output files can be seen in Figure 21 An example of the output file containing All Data is shown in Figure 22 J 19 File Edit View Favorites
59. ce with various types of equipment through site surveys Switches Cisco BTS and Cedar Point MGX DACS routers servers test gear and a variety of fibers e Interacted with outside vendors integrators including Cisco Agilent Motorola Tellabs Sun Cedar Point etc e Acted as Project Planner for test server upgrades for all switch locations around the United States e Developed a nationwide network diagram to demonstrate all current and planned switch locations e Created various organizational tools for documenting unused equipment to prevent budget complications V Comm Telecommunications Engineering Blue Bell PA Network Engineering Co op April 2006 to September 2006 e Assisted engineers in the design process of public switched telephone networks PSTN e Supported engineers in various Switch Interconnection Configuration Plans SICP for Comcast s network e Obtained an understanding of the general principals of telephony from a switching perspective e Gained experience in researching existing network designs As Builts subscriber forecasts and LERG databases e Utilized Microsoft Office and Visio for organizing design data and drawing network design diagrams Honors and Awards Drexel University Dean s Scholar Award 2004 09 Drexel University Dean s List Winter 2004 05 Winter 2006 07 Fall 2007 08 Winter 2007 08 Fall 2008 09 e Pennsylvania Higher Education Assistance Agency PHEAA Academic Excellence Scholar 2004 09
60. creen Save Download new I Always update firmware Show Defaults Load EE Modem XBEE PRO Function Set Version xePos DM XBEE PRO DIGIMESH 900 1820 J E NN Network Delay Slots H E Addressing D SH Serial Number High D SL Serial Number Low D DH Destination Address High D DL Destination Address Low fa COORDINATOR NI Node Identifier NT Node Discovery Backoff Sy Security EE Encryption Enable KY AES Encryption Key 3 Serial Interfacing D BD Baud Rate NB Parity AO Packetization Timeout FT Flow Control Threshold D AP API Enable AO API Options SQ 1 0 Settings D DO 4D0 D100 Configuration Set tead Node Identifier string COM3 9600 8 N 1 FLOW NONE Figure 5 Enter Coordinator Identifier Sleep Configuration 1 Click on the Modem Configuration Tab 2 Click on Sleep Options 3 To set this node to be the coordinator select 1 PREFERRED SLEEP COORDINATOR ENABLE As shown in Figure 6 4 The next thing that needs to be determined is whether the network will be in normal continuous mode or cyclic sleep mode 5 This option is found under SM Sleep Mode e To set the network to normal select 0 e To set the network to Cyclic Sleep select 4 e The COORDINATOR will control the sleep cycles of the other modules in the network e Torun the system as specified select option 4 As shown in Figure 7 6 Set sleep period SP Cyclic Sleep Period e Value is i
61. crew or bolt This is shown in Figure 7 Fasten the XBee Pro interface board down at the top of the enclosure with the use of the Velcro i To do this first attach the Velcro to the back side of the interface board and then attach to the enclosure Place Velcro on the face of the Dylos sensor and then attach it to the bottom of the enclosure Create a loop for the battery on the left side of the enclosure i Take one piece of Velcro with two sides of equal length ii Cut the middle of one of the two sides ili Stick the two parts together making sure that the sticky exposed side of the uncut side is in the middle iv Attach the sticky portion to the enclosure as shown in figure 9 v Place the battery inside the loop and close the loop by Velcroing the sides together In order to ensure no air from the exhaust of the Dylos enters the intake a plastic tube needs to be Velcroed to the Dylos intake 1 9 i Take a plastic tube and cut out a portion the size of the Dylos intake ii Place Velcro on the sides that were cut iii Attach to the outside of the Dylos intake 6 The node should now be complete and can be closed shut and ready for deployment Figure 9 Complete System Node User Manual Part 1 Network Configuration Software Package Part 1 Contents Introduction to Network Configuration Software Installation of X CTU Software Installation of USB Drivers Configuration of Coordinator Module Conf
62. d 4 hours since problem below 14400 seconds timeStampInt 14400 Convert stamp to readable format time_t timeT time_t timeStampInt char time ctime amp timeT Change format from char to string string timeString time int timeStringSize timeString size string timeStringSub timeString substr 0 timeStringSize 1 Week Day Identifier Sun Mon C 10 string weekDayID timeStringSub substr 0 3 Date Identifiers Monday Day Y ear Month Identifier string monthID timeStringSub substr 4 3 if monthID Jan monthID 1 else if monthID Feb monthID 2 else if monthID Mar monthID 3 else if monthID Apr monthID 4 else if monthID May monthID 5 else if monthID Jun monthID 6 else if monthID Jul monthID 7 else if monthID Aug monthID 8 else if monthID Sep monthID 9 else if monthID Oct monthID 10 else if monthID Nov monthID 11 else if monthID Dec monthID 12 else monthID ERROR Day Identifier string dayID timeStringSub substr 8 2 string dayIDFirst timeStringSub substr 8 1 if dayIDFirst 0 dayID timeStringSub substr 9 1 Year Identifier string yearID timeStringSub substr 20 4 Hour Identifier string hourID timeStringSub substr 1 1 2 string hourIDFirst timeStringSub substr 11 1 if hourIDFirst 0 hourID ti
63. d coming technologies such as WiMAX and DVB H to so the company could become more adept in those fields Centralized Odor System Freshman Design e Conducted direct market research amp online research analysis e Presented final report to faculty and fellow students Skills Software Microsoft Office DOORS NetPlan EDX MapInfo Professional Terrain Navigator Pro Maple MATLAB PSpice LaTex writing Programming Languages Basics of Java MATLAB and C Honors and Awards e Drexel Dean s List Fall Winter Spring 04 05 Winter 05 06 Winter 06 07 Fall Winter 07 08 e 2006 2007 Inter Fraternity Council Athlete of the Year e AJ Drexel Scholar June 2004 present e Drexel University Pennoni Honors College Member September 2004 present e Edward J Bloustein Scholar Leadership and Organizational Experience Sigma Phi Epsilon fraternity Chaplain Oversaw standards board ruled on judicial matters and authored new bylaws Internal Relations Chair Oversaw executive board made sure all members were doing their job and set goals for each e board member Community Service Chair Worked with community groups and other student organizations to set up worthwhile community service opportunities Athletic Chair Set up practices and determined teams for each sport Dac Pack Executive Board member Worked with other members to plan trips and events for students Relevant Coursework Calculus I II III Physics 1 II Ill Programmin
64. dity between 40 49 else if humidityInt gt 40 amp amp humidityInt lt 49 conc25 conc25First humidity 100 factor40 concl0 conce25 concLarge goto stop Humidity between 50 59 else if humidityInt gt 50 amp amp humidityInt lt 59 conc25 conc25First humidity 100 factor50 concl0 cone25 concLarge goto stop Humidity between 60 69 else if humidityInt gt 60 amp amp humidityInt lt 69 conc25 conc25First humidity 100 factor60 concl0 conce25 concLarge goto stop Humidity between 70 79 else if humidityInt gt 70 amp amp humidityInt lt 79 conc25 conc25First humidity 100 factor70 concl0 cone25 concLarge goto stop C 17 Humidity between 80 89 else if humidityInt gt 80 amp amp humidityInt lt 89 conc25 conc25First humidity 100 factor80 concl0 conce25 concLarge goto stop Humidity between 90 100 else if humidityInt gt 90 amp amp humidityInt lt 100 conc25 conc25First humidity 100 factor90 conc10 conc25 concLarge goto stop else cout lt lt Error see humidity correction section of code S1 No Rain lt lt endl stop weatherFile close EEEE EEEE EEE EE EEE EEEE EEEE E EEEE CGE E E EEEE EE E CIEE CIR CK Part 5c Error Messages based on Algorithm Output PRR A SGC EEE EEE EEEE EEEE ICCC CRISS CGS E E EEEE EEEE E EEEE EEEE IEC I C
65. documents Location Table Garden 8th and Mercy St HCC Roof 1 Facing 8th St HCC Roof 2 Facing Snyder St Not Assigned Not Assigned Figure 24 Graphical Analysis Output Node Location Names J 21 Warning Report If PM 2 5 gt 35ug m 3 If PM 10 gt 150g m 3 WDay Date Time Node Location Humidity Conditions Conc PM 2 5 Conc PM 10 Message Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri Fri 5 1 2009 1 10 Garden 8th and Mercy St 5 1 2009 1 20 Garden 8th and Mercy St 5 1 2009 1 31 Garden 8th and Mercy St 5 1 2009 1 31 HCC Roof 2 Facing Snyder St 5 1 2009 1 40 HCC Roof 1 Facing 8th St 5 1 2009 1 41 Garden 8th and Mercy St 5 1 2009 1 41 HCC Roof 2 Facing Snyder St 5 1 2009 1 50 HCC Roof 1 Facing 8th St 5 1 2009 1 51 Garden 8th and Mercy St 5 1 2009 1 51 HCC Roof 2 Facing Snyder St 5 1 2009 2 00 Garden 8th and Mercy St 5 1 2009 2 01 HCC Roof 2 Facing Snyder St 5 1 2009 2 01 HCC Roof 1 Facing 8th St 5 1 2009 2 10 HCC Roof 1 Facing 8th St 5 1 2009 2 11 Garden 8th and Mercy St 5 1 2009 2 11 HCC Roof 2 Facing Snyder St 5 1 2009 2 20 HCC Roof 1 Facing 8th St 5 1 2009 2 21 Garden 8th and Mercy St 5 1 2009 2 21 HCC Roof 2 Facing Snyder St 5 1 2009 2 30 Garden 8th and Mercv St SBSSRRRRBRERBERREKEREEER 10 5441 11 5037 11 3356 12 307 14
66. duation June 2009 Cumulative GPA 3 67 Experience Drexel University ACIN Program Camden NJ Modeling and Simulation Co op April 2008 to Present e Developed Diffraction code to estimate loss due to obstructions in terrain e Developed scenarios in JCSS for comparison with COMPOSER e Orchestrated comparison between JCSS and COMPOSER with Lockheed Martin e Authored white paper describing issues with Terrain in JCSS as well as quarterly reports describing progress e Attended annual JCSS conference at the Pentagon L 3 Communications CS East Camden NJ Systems Engineering Co op April to September 2007 e Administrated and assisted engineers with Systems Engineering tools including Doors as well as day to day operations e Received security clearance to work on projects requiring clearance e Tracked day to day usage of Doors by all users to determine if the amount of licenses was adequate e Developed scripts using Perl and DXL languages to assist in day to day operations and maintenance e Updated Processes including training powerpoints tables and Visio flow charts to reflect newest process updates V comm L L C Blue Bell PA RF co op April to September 2006 e Assisted in design implementaion and operation of cellular PCS and DVB H networks e Fine tuned existing wireless networks to provide better service e Prepared coverage plots and documents using MapInfo NetPlan EDX and Terrain Navigator Pro e Researched up an
67. e Note this parameter does not need to be set for synchronized cyclic sleep mode to function COM3 9600 8 N 1 FLOW NONE Figure 6 Enable Preferred Sleep Coordinator J 7 gt X CTU COM3 Remote Configuration PC Settings Range Test Terminal Modem Configuration Modem Parameters and Firmware Parameter View gt Profile r Versions _ Read Write Restore Clear Screen Save Download new I Always update firmware Show Defaults Load versions Modem XBEE PRO Function Set Version XBP09 DM z XBEE PRO DIGIMESH 900 bA 1820 z B MO PWMO duty cycle a B M1 PWM1 duty cycle J D RP RSSI PWM Timer D PR Pull up Resistor Enable Sy AT Command Options B CT AT Command Mode Timeout D GT Guard Times CC Command Sequence Character Diagnostic Commands D VR Firmware Version B HY Hardware Version B 2 Supply Voltage f ER Receive Error Count GD Receive Good Count TR Delivery Failure Count Sleep Commands B 1 S0 Sleep Options D 4 SM Sleep Mode D ST Wake Period D SP Cyclic Sleep Per 1 N A Normal mode is always awake Ct f s YA SP time then wakes for ST 3 time If D7 1 CTS flow control willas pds COM3 9600 8 N 1 FLOW NONE gt X CTU COM3 Remote Configuration PC Settings Range Test Terminal Modem Configuration r Modem Paametersand Firmware p Parameter View p Profile Versions Read wis Restore Clear Screen
68. e Member of Phi Eta Sigma National Honor Society Drexel University Chapter 2004 05 Robert J Falcone 387 Westbourne Dr Broomall PA 19008 3739 robert john falcone drexel edu Education Drexel University Philadelphia PA Bachelor of Science in Electrical Engineering Degree Expected June 2009 Cumulative 3 71 GPA Relevant Coursework Electromagnetic Fields and Waves Introduction to Modulation and Coding Transform Methods and Filtering Electrical Engineering Laboratory Programming for Engineers Engineering Ethics and Professionalism Work Experience Ewingcole Philadelphia PA Electrical Engineering Co op Sports and Entertainment Group April 2007 September 2007 e Assisted in the design of electrical systems for the 1 6 billion New Meadowlands Stadium e Performed load and demand calculations for various electrical products Analyzed panel capacities and designated circuits accordingly e Created layout drawings and aiming patterns based on vendor specifications for Stadium sports lighting Drafted schematic diagrams of Stadium electrical rooms Lockheed Martin Maritime Systems and Sensors Moorestown NJ Co op Technical Senior Ship Electrical Systems April 2006 September 2006 e Created and maintained technical drawings supporting various Naval combat systems e Executed quality control and verification checks of various program electrical drawings e Collaborated on the development of a baseline drawing to be used across several US
69. e is very important because it sets the module to sleep and wake for a programmed amount of time allowing overall power consumption to be very low In this design the module sleeps for 9 minutes and wake for slightly longer than a minute The wake time is set slightly longer than a minute to prevent error because the Dylos takes 60 seconds to transmit the data over a DB 9 cable to the module The sleep mode must also be enabled to allow the voltage regulator circuit to work correctly at each node Power Power Decision Process In order to determine the power consumption of each system component tests were run at 9V using a DC power supply From this power supply a supply current could be obtained for the air sensor and the XBee device The XBee board was found to run at 82 mA in the on mode and 16 mA in sleep mode The Dylos air sensor was found to run at approximately 250 mA in its on mode 128 mA in monitor mode and 76 mA when the sensor was turned off but was still plugged in Battery Power The power supply located at each node in the system is an Impact IMBPD8000 BPD 8000 Rechargeable Battery This device was ideal as it supplies power at a voltage of 9V a requirement of both the Dylos air sensor and the XBee chip The battery has an amp rating of 8 Amp hour resulting in 72 Watt hour for each node in the system After extensive searching and comparison to other batteries on the market this battery was selected due to its lower co
70. ed with the fabrication of a working system for their school ABSTRACT Detection of dangerous particulate matter in ambient air by regulatory groups such as the Environmental Protection Agency EPA has proven to be an expensive and time consuming process The design team has developed a low cost yet effective preliminary screening solution to determine if more comprehensive detection testing is required The system utilizes several laser particle counters LPCs to gather air quality data at different locations each LPC is coupled with a wireless networking device which transmits collected data to a central hub The collected data is passed through a conversion algorithm which converts particle count to concentration so that it can be compared directly to the EPA standards Calibration was performed at the Air Management Service s AMS Lycoming Avenue Testing Facility in Northeast Philadelphia The developed system was compared to the testing solutions utilized by the AMS to determine sensor accuracy and make improvements The solution in place will increase the number of neighborhoods that can be evaluated for particulate matter exposure Currently deployed in South Philadelphia the system conserves valuable time and capital by ensuring that only the areas most at risk are subjected to EPA testing This design process also incorporated high school students from the Science Leadership Academy in Philadelphia as part of The Drexel E
71. ell Range A1 Select Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCell Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results Location 4 Data txt Open FName For Input Access Read As 5 Sep nn Use while loop to copy text file into worksheet While Not EOF 5 Line Input 5 WholeLine If Right WholeLine 1 lt gt Sep Then WholeLine WholeLine amp Sep End If CoINdx SaveColINdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value TempVal Pos NextPos 1 CoINdx CoINdx 1 NextPos InStr Pos WholeLine Sep Wend RowNdx RowNdx 1 Wend Close 5 Format Time Columns C C Select Selection NumberFormat 409 m d yy h mm AM PM C 31 Go to first cell Range A1 Select 13K 2 EE CIC IC EE CIC ICC CK KC IC PREE aK a I I ee aK 2 aK aK a Part 1f Import Location 5 Data txt to sheet Location 5 Data Vf 621 21 1 EI C2 a E E C2 1 CI ICI A 1 EK IC I 1 IC 12 a CI a A A a CA I a E AR Ca a aC a a a Select sheet Sheets Location 5 Data Select Clear current data in sheet Sheets Location 5 Data Cells Clear Begin copying file on first cell Range A1 Select Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCel
72. er thy curiae ek dea seus Ses eed 3 Figure 6 Continuous Data From Southeast Philadelphia System Deployment 7 Figure 7 Winter Term Work Schedule Weeks 1 6 cccccecsseesteceteceeeeeeeeesseenteenaes A 1 Figure 8 Winter Term Work Schedule Weeks 7 11 cccccesccesceeeseeeseeeeeceeeeeeneeenees A 2 Figure 9 Spring Term Work Schedule Weeks 1 6 cccecccesceeeeeseeeneeeeeceseeeeeeetaees A 3 Figure 10 Spring Term Work Schedule Weeks 7 11 c ccc eeceeeceeeseeeseeeeeceteeeeeeennees A 4 Figure 11 System Design Budget Comparison ccccccescceeseeesseceseceteceseeeeseecsaeeneenees B 1 Figure 12 Industrial Budget Comparisotis 3csedoccsdaxisassesecajiaislaceasiseasiaielaeaduaeniees B 2 Figure 13 Comparison of AMS Data to ECE 19 Uncorrected Data ssesessseessesee D 1 Figure 14 Humidity Data During Deployment ssnsseseseseessesseseesseesresresseeseserssessesse D 1 Figure 15 Comparison of AMS Data to ECE 19 Corrected Data cc eeeeseeeeeeees D 1 Figure 16 Comparison of AMS Data to ECE 19 Corrected Data Sensor 2 D 2 Figure 17 Comparison of Daily Averages Filter Method to AMS and ECE 19 Continuous Aletta t arate ete heh ai Cee aN ete ae a a a Na ae alg D 2 Figure 18 Size Comparison of System Node Size to FEM Monitor 0 0 eeeeeeeereenee D 3 Figure 19 AMS Filter Weighing System sx ssccicess sais nssandeschs cade adeadsucta caebievs casaeespaclonioss D 4 Figure 20 Sy
73. es in utilizing the XBee Pro DigiMesh 900 development kit The main advantage is that this kit contains all of the necessary hardware needed to begin deploying a wireless network including the XBee Pro modules interface boards antennas and cables To interface a wireless module to the Dylos DC1100 sensor an RS 232 interface board is used in combination with the XBee Pro Module The mounting of the module is shown in Figure 8 Appendix H 16 The serial interfaces of both the node module and the Dylos DC1100 Pro sensor allow the transmission of data between the two devices through a standard DB 9 cable Once the data is transmitted over the DB 9 cable the module is able to transmit the information wirelessly to a coordinator module located at the base station The coordinator module is similar to the node modules but contains a USB interface board to connect directly to an Asus Eee PC which then collects and stores all of the data from each sensor Software Another advantage of utilizing the XBee Pro DigiMesh 900 development kit is the use of free configuration and testing software The software included with the kit is called X CTU By utilizing this software each module can be programmed with several different configuration attributes 16 The main configuration screen for each modem or module is shown in Figure 9 Appendix H The two main configurations used in this design are the cyclic sleep mode and the mesh capability The sleep mod
74. es less than 2 5 um in size for certain lengths of time 1 2 3 Studies evaluating the link between particulate matter and cardiopulmonary mortality have determined that approximately 50 000 to 100 000 Americans die each year as a result of exposure to these dangerous pollutants 4 In the Philadelphia region particulate matter monitoring is funded by the Environmental Protection Agency EPA and operated by the Philadelphia Air Management Service AMS 5 The AMS s main system of monitoring is based on the EPA s Federal Reference Method this is the method of passing a sample of air through a filter manually removing the filter weighing it and determining the concentration of hazardous particulate matter in the given sample in ug m see Table 1 This method gives extremely accurate results but is both expensive and time consuming Additionally the filters yield only one reading per day as opposed to many readings complete with time stamps which would allow for the investigation of any hazardous event that would cause a spike in particulate matter The cost for a technician to drive to replace and transport a filter can cost approximately 100 200 each time 6 In addition the cost to set up a regional weighing lab is approximately 25 000 to 50 000 These costs are incurred by the individual states which draw from a shared pool of funds totaling approximately 40 million each year 7 Wireless Module System Diagram
75. fficient funds Collaboration with the Clean Air Council has allowed the design team the opportunity to deploy the system at the Houston Community Center 2029 South g Street in Southeast Philadelphia The design team worked hand in hand with the CAC to ensure that all appropriate deployment procedures were followed Though the system has been deployed in this specific area it has been developed for deployment at any location making it a viable option for a number of locales looking for a cursory determination of the quality of their air Through the Drexel EPICS Program the team has been able to truly influence the educational lives of young engineering students The Science Leadership Academy students were not only extremely intelligent but also hungry for knowledge possessed a strong desire to help impact their community s environment The design team was pleased to witness a noticeable increase in the students comprehension of the engineering process and admired their newfound interest in Drexel University Photos from the design team s work with the high school students can be seen in Figures 22 23 Appendix F Constraints Economic Cost effectiveness was a key goal in design and fabrication of the system as it attempted to help solve the problem of high costs of current particulate matter testing Thus in an attempt to keep costs down some system components for example the voltage regulator circuit were fabricated by hand
76. ftware er eas se Documentation Software Product Families XBee Modules XBee Pro 900 and DigiMesh 900 XBee Modules DigiMesh 900 Data Sheet amp Documentation XBee product page Install X CTU EXE gt Select a document internet link or application Figure 2 Installation of X CTU Software C Installation of USB Drivers In order to utilize the supplied USB interface board drivers must first be installed by using the supplied Hardware and Software Setup CD The two drivers that need to be installed on the PC are a USB driver and a virtual COM port driver The COM port driver allows the USB port to perform like a physical COM port Installation Details 1 First insert the CD is in the drive before connecting the USB interface board 2 Connect the USB interface board and wait for the Found New Hardware Wizard Plug and play device 3 Click Continue Anyway when the alert box shows up 4 This will install the USB Driver 5 Follow the steps to install the virtual COM port driver which will show up after the first driver is installed 6 Click Continue Anyway when the alert box shows up 7 The necessary drivers will now be installed on the system J 3 D Configuration of Coordinator Module After completing the software driver installations the X CTU software can be used to perform network configuration This process begins with the configuration of the Coordinator module th
77. g 4 28 5 12 15 z 5 1 Development of C Package 3 30 5 12 45 5 2 Development of MATLAB Code 3 30 5 12 45 6 1 Final Report Preparation 4 20 5 13 24 6 2 Final Presentation Preparation 4 27 5 22 26 KEIT a Figure 10 Spring Term Work Schedule Weeks 7 11 Legend All Rob Falcone Anthony McClellan Ed Ostapowicz Mark Uva A 4 Appendix B Budget Proposed Budget Pat COTY Benefits Price Ea Total _ Dylos DC1100 Pro 305 00 1 220 00 E a A tual an Nodes and Base Station 200 00 1 200 00 Interface Cables Serial USB Etc 10 00 Single Board Computers DA 150 00 150 00 Batteries Batteries Tel s6000 360 00 Antenna o el soo 180 00 Cellular Connection Phone USB Modem or PCMCIA Card 1 150 00 150 00 Per Month 17 39 99 199 15 Weatherproof Enclosures EIEE 50 00 300 00 G so 180 00 60 00 Figure 11 System Design Budget Comparison Actual Budget Part OTY Price Ea Total __ Custom Caitraton 6 san4s4 51 2470 Custom Calibration 5 304 94 1 524 70 2 PO Development Kit 3 216 14 648 42 USB to Serial Cable 10 00 10 00 Asus Eee PC 2 299 00 598 00 51 84 516 40 Node Enclosures 33 00 190 0 Base Station Enclosure 45 00 45 0 ee 50 Voltage Regulator Circuit Components oo son 000 Comparison Cost details Higher cost than proposed Lower cost than proposed Lo
78. g for Engineers Java Fundamentals of Intellegent Systems Electric Circuits Energy Il Systems II Linear Modeling Vector Analysis Transform Methods Digital Signal Processing Electronic Devices Analog Electronics Intro to Modulation and Coding Electromagnetic Fields and Waves Wireless Communications Wireless and Optical Electronics ECE Lab L I II IV Communications l II Deterministic Signal Processing Statistical Signal Processing Edward W Ostapowicz 3867 Marsh Rd Garnet Valley PA 19061 610 494 2664 edward w ostapowicz drexel edu Education Drexel University Philadelphia PA Bachelor of Science in Electrical Engineering Anticipated Graduation June 2009 GPA 3 73 Relevant Coursework Fundamentals in Intelligent Systems Basic Java Programming Electrical Engineering Lab Il Ill IV Fundamentals in Signals and Systems Electronic Devices Motor Control Principles Energy Management Principles Power Electronics Power Systems Systems and Control Computer Skills e Basic Java e Basic AutoCAD e Basic SKM Power Tools e Microsoft Office e Basic Maple e Microsoft Operating Systems Basic MATLAB e Basic Microstation Work Experience Sunoco Inc Philadelphia PA Instrument Electrical Reliability Engineer Co op October 2007 to April 2008 Updated database using EMPAC software e Designed relay rack mount using Microsoft Visio e Assisted in design of outdoor Class A Div II relay mount e Assisted i
79. ge Regulator Circuit 4 5 Test Power Consumption With System 5 1 Research Algorithm Further 2 Finalize Algorithm 6 1 Work On Progress Repon Pe aT 6 2 WorkOnPresenition E a TT TT PT TT E Figure 8 Winter Term Work Schedule Weeks 7 11 Legend All Rob Falcone Anthony McClellan Ed Ostapowicz Mark Uva A 2 Ttem 2l Task Name Validation with AMS Start Date 3 30 End Date 5 12 Time Month of April 2009 Duration Days Lif 2l 3 al 5 e 7 8 9 iol 1l 12 13 14 15 16 17 18 19 20 21 22 23 24 25 as 3 1 Finelixation of Algorithm 3 30 5 12 45 4 1 South Philly Deployment T esting Development of C Package Development of MATLAB Code Final Report Preparation 4 28 3 30 3 30 4 20 5 12 5 12 5 12 5 13 kh hase seis ass aa se Sass 45 45 Final Presentation Preparation Weekly Meetings 4 27 444 5 22 5 22 24 A l A l i i E l eee eee ee Figure 9 Spring Term Work Schedule Weeks 1 6 Legend All Rob Falcone Anthony McClellan Ed Ostapowicz Mark Uva A 3 Month of May Week 5 Week 7 Week 8 Week 5 26 27 28 29 30 al 2 3 al s el 7 sl o 1o iif 12 13 14 15 16 17 18 19 20 21 22 23 4 513 7 _2 Weekly Meetings with advisors 4 3 5 15 F Validation with AMS 3 30 5 12 45 Finalixation of Algorithm 3 30 5 12 45 4 1 South Philly Deployment Testin
80. gth indicator Feicent 100 Good wr eaat onosawz Bad amas p Figure 11 Check box for RSSI indicator SS CTU CONF About pee BARCDIICHOINLAUD pawn BARCDIFORTIKLINO ee BABCDIPHLJELINO pE BABCI IICHLJKLINO i p vw BABCDIICHIIELIMO p wew BABCDIIGHOINEMND BADCDIFGRITIKLAND pee BARCDIICHOINLAUD 0123456763 z471 RABCDIFOHOIFLIOD 0123456793 e BABCDIICHLIRLAND 0129456789 pte BABCDIFCHLIRLAMD i p wr BABCDSIGHIINLIOND BABCDIIGHOINLAND gt BARCDIFCHIINLAND pee BARCDIICHTINLIND Trent Recewe Fans O507 COM SSD0EN 1 FLOW NDNE Figure 12 Range Test in Action J 11 User Manual Part 2 Algorithm Implementation Software Package Part 2 Contents A Phase 1 Data Collection 1 RealTerm B Phase 2 Data Conversion and Analysis 1 Import Weather Data 2 Data Extraction and Algorithm Implementation 3 Graphical Analysis Overview The algorithm implementation software package is broken up into two phases The first phase requires capturing and saving the data during system deployment After the data has been collected the data enters the second phase of the software process In the second phase the captured data is extracted converted to the EPA scale ug m and graphed for analysis A block diagram of this process is shown in Figure 13 Import Weather Data Purpose Captures weather information for each day input by the user Hum
81. icated the crucial voltage regulator circuits and was also the team s main point of contact to the Clean Air Council 11 Robert Falcone derived the initial particle count to conversion algorithm used within the team s data analysis software package He also coordinated the written reports and oral presentations and led meetings with the Science Leadership Academy students Conclusion Considerable progress has been made on the preliminary screening system for particulate matter in ambient air since work began in the fall of 2008 While all stated deliverables have been met in the first fabricated version of the system there is still room for improvement based on existing constraints Suggestions for future work include e Further reduce power consumption to extend battery life e Perfect mesh networking structure of Xbee hardware to increase system range e Improve data conversion algorithm based on further testing e Utilize creative marketing in conjunction with test results to obtain deployment permission from a greater number of residents within a given test area The design team has learned a great deal about the issue of particulate matter in ambient air the problems inherent in the current testing solutions and about the methods of comparison to the accepted standards Furthermore knowledge of the engineering process has been gathered firsthand and will undoubtedly provide useful to the members of the design team in the futu
82. idity and rain conditions are captured for the algorithm implementation Output Comma Separated Values File CSV Code type VBA Creator ECE Team 19 Weather Source http weather myphl17 com PHASE 1 RealTerm Description Serial terminal Purpose Captures data received wirelessly by XBee Coordinator connected to any COM port on the data collecting PC Output Text file in ascii Creator Open Source http realterm sourceforge net PHASE 1 Data Capture Phase PHASE 2 Data Extraction and Algorithm Implementation Inputs RealTerm output file txt and imported weather data csv Purpose Utilizes inputs to perform the detailed particle count to mass conversion The data is then reformatted into several text files The text files included are warning report averages and mass concentrations for each node location Output Various text file Code type C Creator ECE Team 19 Detailed Analysis Inputs The various text files output from data extraction and algorithm implementation Purpose Auto imports all text files into easily readable Excel spreadsheets The results for each node location are also displayed through graphs Code type VBA Creator ECE Team 19 PHASE 2 Data Extraction Algorithm Implementation and Graphical Analysis Figure 13 Software Block Diagram A PHASE 1 Data Collection Program A Realterm Reallerm Output Figure 14
83. iguration of System Node Module Test Network Range mt et gt A Introduction to Network Configuration Software This manual assumes that the system hardware is complete as shown in Figure 1 If this is not the case please refer to the Engineering Manual Hardware Development for assistance This software manual provides the details involved in properly installing the XBee Pro development kit supplied software and drivers After installation is completed this manual further describes the details in configuring a successful network J 1 Figure 1 Node System Setup B Installation of X CTU Software X CTU is the network configuration software that is supplied with the XBee Pro DigiMesh 900 development kit The software is contained on a CD labeled Hardware and Software Setup This software must be installed to perform network configuration Installation Details 1 Insert the supplied CD and wait for the set up screen to appear 2 Click on Modules Sensors and Adaptors Documentation Software 3 Then click XBee Modules 4 Then click on DigiMesh 900 5 Then click on install X CTU Software as shown in Figure 2 6 Follow the onscreen instructions to install the X CTU Software gt s Digi Drop in Networking om Drop in Networking 75000564 sO Hardware amp Software Setup Gteway Host Enterprise Documentation Software Gatewa Modules Sensors amp Adapters Docume wart Documentation So
84. igure 7 Voltage Regulator Circuit Attachment to XBee Interface Board 6 Power Source 9V Battery The power source chosen for this system is the Impact universal Li Ion rechargeable battery as shown in Figure 8 This battery was chosen due to its high capacity to cost ratio An alternative battery may be selected if it supplies an output of 9V Device Details Supplier B amp H bhphotovideo com Part Number BPD 8000 Capacity 8000mAh Output Voltage 9 10V Connections i Out connection between the battery and the voltage regulator circuit with supplied cable ii In used to charge the battery with supplied AC adapter Figure 8 Impact 9V Battery 1 8 7 Completion of System Node Once all of the parts are assembled they must be placed securely within a weatherproof enclosure The Primex P1000 enclosure has been used in this system due to its large depth and weatherproof hinges A large depth is needed to completely enclose the Dylos sensor s irregular shape The completed system is shown in Figure 9 below a Overview of System Node Components Qe Dylos DC1100 Pro Laser Particle Counter XBee Pro DigiMesh 900 Module and Interface Boards Null Modem Serial Cable RS 232 Interface Voltage Regulator Circuit Power Source 9V Battery b Mounting Components in Enclosure See Figure 9 l 5 Secure the voltage regulator circuit to the XBee Pro serial interface board by using a small s
85. ikes seen at all three locations on May 1 2009 are likely due to the high humidity seen during that particular day minimum humidity 78 maximum humidity 100 and average humidity 90 However evaluation of more data would be required to determine whether these spikes are repeated at the same time each week Based on the data collected from the Southeast Philadelphia the design team would not recommend that more expensive and comprehensive testing take place It would also recommend an investigation into any possible events within the neighborhood that could have caused the elevated concentration levels seen on May 1 2009 However it is important to note that the collection of more data using the preliminary screening system can only improve on the accuracy of any recommendations Societal Ethical and Environmental Impacts Each day more hazardous materials are finding their way into ambient air with fewer state dollars available to provide relief 6 The result by simply breathing more and more people will face adverse health effects in the short and long term These health effects are not trivial and range from brain and lung damage to death With less money available to help test for particulate matter the developed system will ensure that the high cost of extensive testing is endured only when absolutely necessary furthermore more areas can be screened to determine if they are at risk as opposed to testing them at all due to insu
86. ined that orienting the enclosure with its vents facing down was the only feasible option This constraint also limited where the Dylos sensor could be placed within the enclosure as to maximize the amount of air it can capture while still shielding the wireless boards from moisture Ethical Health and Safety Because protection of the public was a driving factor in the system s development ethical health and safety provided motivation as opposed to constraint At the same time the team does recognize the need to abide by the ethical reporting code in the collection of its data findings should be presented to the appropriate governing bodies rather than taking results directly to the public forum Social The role of the public in the system s success can not be understated While some government owned buildings are available to potentially host system components public participation would allow a far greater area to potentially be screened In addition to public participation the design team needed to keep in mind the possibility of individuals tampering with system components Thus the decision to hang components only in inconspicuous locations at heights out of reach of pedestrians was made Political The design team learned that it must operate under the watchful eye of governing bodies such as the Clean Air Council and Air Management Service These organizations are the current authorities on public environmental issues thus
87. l Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results Location 5 Data txt Open FName For Input Access Read As 6 Sep NER nan Use while loop to copy text file into worksheet While Not EOF 6 Line Input 6 WholeLine If Right WholeLine 1 lt gt Sep Then WholeLine WholeLine amp Sep End If CoINdx SaveColINdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value Temp Val Pos NextPos 1 CoINdx CoINdx 1 NextPos InStr Pos WholeLine Sep Wend RowNdx RowNdx 1 Wend Close 6 Format Time Columns C C Select Selection NumberFormat 409 m d yy h mm AM PM Go to first cell Range A1 Select UK KC IC ICICI CE CSIC ICC 2 CK ICCC KK KIC a I ee aK 2 aK aK ak Part 1g Import Averages txt to sheet Daily Averages USK 2 2 I I CC IC EE CSIC IC CE 2 KIC IC KK aK I a a ee ok aK oi ok aK ak C 32 Select sheet Sheets Daily Averages Select Clear current data in sheet Sheets Daily Averages Cells Clear Begin copying file on first cell Range A1 Select Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCell Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results Averages txt Open FName For Input Access Read As 7
88. le Once the data is sent to the networking component it is then sent wirelessly to the coordinator at the base station This process is completed by each of the sensors and the data is stored and then analyzed at a remote location Power Consumption G 1 Due to the high current draw of the Dylos sensor 250A at 9V it was necessary to develop a voltage regulator circuit This circuit would allow the sensor to be powered on while a reading was being taken but powered off when a reading was not necessary More specifically a reading would be taken 6 times per hour for a total of 6 minutes per hour and the device would be turned off for 54 minutes per hour This would allow a significant amount of readings to be taken but at the same time conserving battery power For more information regarding this circuit please see the Power section Networking Network Decision Process The decision to use a Xbee standard product was made due to its low power consumption when compared to Wi Fi products the second networking option evaluated for this system Two Xbee options considered were the 2 4 GHz range or the 900 MHz range devices The Digi XBee Pro DigiMesh 900 series utilized 900 MHz and was chosen mainly due to the low cost of the development kits In addition to its low cost the XBee Pro demonstrated an extended range improved sleep modes and mesh networking in comparison to the original XBee 15 Hardware There are many advantag
89. led by clicking Download New Versions 3 Click Read to display the parameter as shown in Figure 4 4 Module Identification e Under Addressing select NI Node Identifier e Enter COORDINATOR as shown in Figure 5 A J 5 gt X CTU COM3 Remote Configuration PC Settings Range Test Terminal Modem Configuration m Modem Parameters and Firmware Parameter View m Profile Versions Write Restore Clear Screen Save Download new J Always update firmware Show Defaults Load versions Modem XBEE PRO Function Set Version XBPOS DM XBEE PRO DIGIMESH 900 7 1820 5 4 Networking A D ID Modem VID a HP Hopping Channel D MA Mesh Retries D NH Network Hops D NG Network RREQ Retries D NN Network Delay Slots Sy Addressing D SH Serial Number High D SL Serial Number Low D DH Destination Address High D DL Destination Address Low D NI Node Identifier D NT Node Discovery Backoff Sg Security EE Encryption Enable D KY AES Encryption Key Serial Interfacing BD Baud Rate D NB Parity Failed to enter command mode Unable to read Version ATYR Read parameters Failed COM3 9600 8 N 1 FLOW NONE Figure 4 XBee Pro Module Parameters gt X CTU COM3 Remote Configuration PC Settings Range Test Terminal Modem Configuration m Modem Parameters and Firmware Parameter View gt j Profle p Versions Read Write Restore Clear S
90. lt lt setw 20 lt lt left lt lt largeAverageLoc2 lt lt lt lt setw 20 lt lt left lt lt messageLoc2 lt lt endl countLoc2Cycle countLoc2Data 0 sumSmallLoc2 0 sumLargeLoc2 0 i Average Calculations countLoc2Data Count total entries for average sumSmallLoc2 conce25 Sum Concentration of 2 5 sumLargeLoc2 conc 10 Sum Concentration of 10 Apply previous variables for average calculation prevYear2 yearID prevMonth2 monthID prevDay2 dayID Location 3 else if locationNumber 3 outputLoc3 lt lt setw 6 lt lt left lt lt Loc3Index lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 16 lt lt left lt lt concatDateTime lt lt lt lt setw 5 lt lt left lt lt nodeldentifier lt lt lt lt setw 35 lt lt left lt lt location lt lt C 22 entries sumSmallLoc3 countLoc3 Data 1 sumLargeLoc3 countLoc3Data 1 150 high lt lt wo lt lt setw 12 lt lt left lt lt humidity lt lt lt lt setw 11 lt lt left lt lt condition lt lt lt lt setw 12 lt lt left lt lt smallCountInt lt lt lt lt setw 12 lt lt left lt lt largeCountInt lt lt lt lt setw 13 lt lt left lt lt conc25 Zeeu lt lt setw 12 lt lt left lt lt conc10 See lt lt setw 7 lt lt left lt lt message lt lt endl L
91. meStringSub substr 12 1 string hourldent hourID Convert hour ID to integer int hourIDInt hourIDInt atoi hourID c_str Minute Identifier C 11 string minuteID timeStringSub substr 14 2 RBBB A ECCI CCCI ICCC E EEE EEE EEEE E EEEE EEEE EEEE E E E Part 3b Extraction of Node Identifier PRB B EEE EEE EEEE EEEE EEEE EE IC CGC EE E EEE EEE EE EEE E ICCC ICICI ICI Extract Node Identifier find Capture 2 character identifier after int secondMarkerLoc line find string nodeldentifier line substr secondMarkerLoc 3 Location of int thirdMarkerLoc line find p firstMarkerLoc 1 Location of second comma int fourthMarkerLoc line find firstMarkerLoc 1 RBBB A CIC C EEE EEEE EEEE E EEE E E EEEE EE EEE E EE ICCC CICERO Part 3c Extraction of Particle Counts ek EEEE EEEE EEEE EEEE E E EE EEEE E E E E E EE EEE EEE E E EEEE EEE EEEE E E E E Extract Small Particle Count PM 2 5 istringstream smallCount string smallCount line substr thirdMarkerLoc 2 fourthMarkerLoc thirdMarkerLoc Extract Large Particle Count PM 10 int lineSize line size Size of entire line string largeCount line substr fourthMarkerLoc 1 lineSize fourthMarkerLoc 1 Particle Count Convert from string to integer int smallCountInt largeCountInt smallCountInt atoi smallCount c_str largeCountInt atoi largeCount c_str PRR R ECCI R EE
92. n Phones USB Modem or PCMCIA Card 150 00 Single Board Computers Po stsooof 150 00 sus Eeerc 2 S sssoof 8598 00 i S a Circuit Components 10 5 00 50 00 P Sool Serial Cables 60 00 Serial to USB ooo 10 00 10 00 Miscellaneous Miscellaneous Software Service Software Service Laptops Windows xP 4 1500 00 6 000 00 Laptops windows xP 4 ssoooof 36 00000 Internet Internet 40 mo for 9 months 120 00 j f j for 9 months Development Software MATLAB license 500 500 00 CTC AT 0 TT Cellular Service Data 39 99 mo 6 months i 199 20 Tota l S 95 080 00 93 399 20 85 639 20 274 118 40 a Total 90 849 88 92 999 16 85 440 00 269 289 04 Figure 12 Industrial Budget Comparison B 2 Appendix C Particle Count to Mass Conversion Algorithm Import Weather Data Code Type VBA Macro Sub Import_Weather Description This program captures all of the weather information needed for the algorithm Humidity and rain conditions The location of the network and start end dates are input on the user input tab Updated 5 12 09 Creator Anthony McClellan Update User name for each computer Dim UserName As String UserName Anthony Capture all Date Location information entered by the user Sheets User Inputs Select Start Date Dim BeginYear As Integer BeginYear Cells 13 D Value Dim Begin
93. n increments of 10 ms e Enter value in HEX format e For 8 min and 55 sec enter DOFC 7 Set Wake Period ST Wake Period e Value is in increments of 1 ms e Enter value in HEX format e For 1 min and 5 sec enter FDE8 8 The parameters should now look like Figure 8 9 Click Write to save these parameters to the XBee Pro Module gt X CTU COM3 Remote Configuration PC Settings Range Test Terminal Modem Configuration Modem Parameters and Firmware Parameter View y Profile Versions Always update firmware Show Defaults Load versions Modem XBEE PRO Function Set Version XBPO9 DM w XBEE PRO DIGIMESH 900 y 1820 v B MO PWMO duty cycle a D M1 PWM1 duty cycle f RP ASSI PWM Timer PR Pull up Resistor Enable SQ AT Command Options D CT AT Command Mode Timeout D GT Guard Times D CC Command Sequence Character 4 Diagnostic Commands D VR Firmware Version D HV Hardware Version D 2 Supply Voltage ER Receive Error Count GD Receive Good Count D TR Delivery Failure Count Sq Sleep Commands SO Sleep Options TERT aes eC li SM Sleep Mode 9 PREFERRED SLEEP COORDINATOR DISABLE ST Wake Period 1 PREFERRED SLEEP COORDINATOR ENABLE D SP Cyclic Sleep Period Pref Sleep Coordinator will cause a Normal node to act as Sleep Coordinator in preference to any sleeper node and will cause a sleeper node to become a Sleep Coordinator if it misses a single synch messag
94. n visual inspections of electrical equipment in the refinery Attended many vendor sessions to learn about new equipment KlingStubbins Philadelphia PA Electrical Engineering Intern September 2006 to April 2007 e Assisted in buildings short circuit analyses e Assisted in CADD design of buildings electrical systems Traveled to Ft Lee VA for on site surveying Coordinated group of three to finish CADD drawings Drexel University Philadelphia PA Aide Engineer Lab Tech October 2005 to April 2006 Prepared laboratories for daily classes e Tested devices used in the labs e Created supplemental components for particular labs e Maintained laboratory equipment e Controlled order within the laboratories Eastern Electric Mantua NJ Blueprint Analyst June to August 2000 and 2001 Did the take offs on blueprints for various future projects e Formulated organized system of double checking the take offs e Interpreted different symbols from table to drawing Tabulated amounts of light fixtures receptacles and other equipment Managed time effectively to meet deadlines e Reported final numbers to supervisor Honors and Awards e Freshman Design Project Published 2006 e Treasurer of the DAC Pack 2007 2008 Intramural Campus Team of the Year Co Captain 2005 Dean s List 2005 2007 e Perfect Attendance 2000 2004 e Neumann Scholarship High School and College A J Drexel Scholarship College e Football St
95. nd Settings Anthony Desktop C Results All Data txt Header outputAlIData Title lt lt setw 9 lt lt left lt lt All Data lt lt endl lt lt endl Header lt lt setw 6 lt lt left lt lt Index lt lt setw 6 lt lt left lt lt WDay lt lt setw 17 lt lt left lt lt Date Time lt lt setw 6 lt lt left lt lt Node lt lt setw 36 lt lt left lt lt Node Location lt lt setw 13 lt lt left lt lt Humidity lt lt setw 12 lt lt left lt lt Conditions lt lt setw 13 lt lt left lt lt Small Count lt lt setw 13 lt lt left lt lt Large Count lt lt setw 14 lt lt left lt lt Conc PM 2 5 lt lt setw 13 lt lt left lt lt Conc PM 10 lt lt setw 8 lt lt left lt lt Message lt lt endl Separate Output Files Location 1 ofstream outputLoc1 outputLocl open C Documents and Settings Anthony Desktop C Results Location 1 Data txt outputLoc1 Title lt lt setw 16 lt lt left lt lt Location 1 Data lt lt endl lt lt endl Header Header C 5 lt lt setw 6 lt lt left lt lt Index lt lt setw 6 lt lt left lt lt WDay lt lt setw 17 lt lt left lt lt Date Time lt lt setw 6 lt lt left lt lt Node lt lt setw 36 lt lt left lt lt Node Location lt lt setw 13 lt lt left lt lt Humidity lt lt setw 12 lt
96. ngineering Projects in Community Service Program il Table of Contents Introduction and Problem Desen prion wsenccosesinsaektectvcan wddeikahamieiaue adonttih aiuadaaddevaecdines 1 Final WSS IO ess a mies sacs aca iecs tea dees Aca seig data a Peta ia dh aa Mee Saco a E E E Accessed eked 2 AMS Test Deployment Calibration and Validation csceseecceeeceseeeeeeeeeeeceseceaeeneeees 4 Particle Count to Concentration Conversion Al gorithm cccccccsceeceeeseeeseeeeteeeteeeees 6 South Philadelphia Deployment cs i sscctee sessivsssdaidecusecortsbesccevsesvsseous gacta ons caveagnicaebsecceess 7 Societal Ethical and Environmental Impacts ss sessssessesseessesseseesseesesresseesseseessresseseess 9 Constraints ae a a a ea a eta Dace Ei 9 Timelines e elects a nes ods e a a E E EA Y 11 B dg ti erreinu ara rena ial a a se a a Aa E E a Ea aS 11 CONC TS LOT ngn aaa aa aei nae eae eae 12 REST TNC OS otras Sica apie ctie cc cute ug te aca e ack nee g nestees tao atone dete eae eed eea eae 13 Appendix A Work Schedule ncen co eSahes csuaes sel tuad arads a A 1 Appendix B B dget serenon e cade a E E OAA E Aas B 1 Appendix C Particle Count to Mass Conversion Algorithm s essseseessssesesesseseeee C 1 Import Weather Data ices 2 cack caked ii E EEE EE E E R C 1 Data Extraction and Algorithm Application ccccccecsseceseceseceeeeeeeeesteceeeeeeeeeaeees C 3 Graphical Analysis xiccai deri sa aria
97. oc3Index Increase to next row value count data if Loc3Index gt 2 if prevDay3 dayID i double smallAverageLoc3 double largeAverageLoc3 Warning messages string messageLoc3 Warn 3 both PM2 5 and PM10 over thresholds if smallAverageLoc3 gt 35 amp amp largeAverageLoc3 gt messageLoc3 Warning PM 2 5 and PM 10 too Warn 1 PM2 5 over 35ug m 3 elseif smallAverageLoc3 gt 35 messageLoc3 Warning PM 2 5 too high Warn 2 PM10 over 150ug m 3 else if largeAverageLoc3 gt 150 messageLoc3 Warning PM 10 too high No Warning else messageLoc3 averages lt lt setw 35 lt lt left lt lt location SANN lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 5 lt lt left lt lt yearID gAn lt lt setw 6 lt lt left lt lt monthID SaN lt lt setw 4 lt lt left lt lt dayID lt a lt lt setw 20 lt lt left lt lt smallAverageLoc3 lt lt setw 20 lt lt left lt lt largeAverageLoc3 lt lt lt lt setw 20 lt lt left lt lt messageLoc3 lt lt endl countLoc3Cyclet countLoc3 Data 0 sumSmallLoc3 0 sumLargeLoc3 0 Average Calculations countLoc3Data Count total entries for average C 23 data entries sumSmallLoc4 countLoc4Data 1 sumLargeLoc4 countLoc4Data 1 150 high sumSmallLoc3 conc25 Sum Concentration of 2 5 sumLargeLoc3 conc10 Sum Concentration
98. of 10 Apply previous variables for average calculation prevYear3 yearID prevMonth3 monthID prevDay3 dayID Location 4 else if locationNumber 4 outputLoc4 lt lt setw 6 lt lt left lt lt Loc4Index lt lt lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 16 lt lt left lt lt concatDateTime lt lt lt lt setw 5 lt lt left lt lt nodeldentifier lt lt lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 12 lt lt left lt lt humidity lt lt lt lt setw 11 lt lt left lt lt condition lt lt lt lt setw 12 lt lt left lt lt smallCountInt lt lt lt lt setw 12 lt lt left lt lt largeCountInt lt lt lt lt setw 13 lt lt left lt lt conc25 KAUT lt lt setw 12 lt lt left lt lt conc10 lt ANT lt lt setw 7 lt lt left lt lt message lt lt endl Loc4Index Increase to next row value count if Loc4Index gt 2 if prevDay4 dayID double smallAverageLoc4 double largeAverageLoc4 Warning messages string messageLoc4 Warn 3 both PM2 5 and PM10 over thresholds if smallAverageLoc4 gt 35 amp amp largeAverageLoc4 gt messageLoc4 Warning PM 2 5 and PM 10 too Warn 1 PM2 5 over 35ug m 3 elseif smallAverageLoc4 gt 35 messageLoc4 Warning PM 2 5 too high Warn 2 PM10 over 150ug m 3 else if largeAverage
99. ons necessary for success The base station was deployed at the Houston Community Center located on 8 and Snyder Streets in Southeast Philadelphia see Figure 20 Appendix E Because no power source was available on the building s roof the unit was secured inside an office on the third floor As a result the nodes could not maintain a communication link several blocks away and had to be deployed closer to the base station The system was deployed and collected data from April 28 to May 4 2009 The first node Location 1 was deployed in a garden facility between 7 and 8 Street on Mercy Street This location was excellent as fencing was available the fencing not only provided a structure from which the node could hang but also kept pedestrians away from the unit Two additional nodes were deployed on the roof of the Houston Community Center One node Location 2 faced 8 Street the other Location 3 faced Snyder Street These units were hung approximately 30 feet from one another and overlooked high traffic urban streets Figure XX Appendix E Figure 6 shows the final system results As previously described the EPA standard specifies that the average concentration for a particular day should not exceed 35 ug m As seen in Table 4 this condition is met for each day of the deployment period om Continuous Data From Southeast Philadelphia Deployment 8 Concentration ug m wa ae bn se i el h aa
100. or Import Weather Data Program Output Files The outputs of this program are numerous comma separated value csv files A file is created for each day and is saved in the Weather Output folder on the desktop An example of the output file is shown in Figure 20 I Weather 2009 5 5 Notepad e xX File Edit Format View Help Time EDT Temperature Dew Point Humidity Pressure Visibility wind Direction wind Speed Gust Speed Precipitation Events Conditions a 0 05 51 8 F 50 0 F 94 30 05 in 5 0 miles NNE 16 1 mph 0 00 in Light Drizzle 1 F 50 0 F 96 30 06 in 7 0 miles NNE 11 5 mph 0 00 in Overcast 8 F 50 0 F 94 30 06 in 7 0 miles NE 10 4 mph N A Overcast 0 F 50 0 F 93 30 05 in 9 0 miles NE 11 5 mph N A Overcast 1 F 51 1 F 100 30 05 in 3 0 miles NE 8 1 mph N A Overcast 8 F 51 8 F 100 30 05 in 2 5 miles NNE 9 2 mph N A Overcast 8 F 51 8 F 100 30 05 in 1 5 miles NNE 9 2 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 1 5 miles NNE 9 2 mph 0 00 in Light Drizzle 8 F 51 8 F 100 30 05 in 1 2 miles NNE 10 4 mph 0 01 in Light Drizzle 8 F 51 8 F 100 30 05 in 1 2 miles NNE 9 2 mph 0 01 in Lig t Drizzle 1 F 51 1 F 100 30 05 in 1 2 miles NNE 11 5 mph 0 01 in Light Drizzle 8 F 51 8 F 100 30 05 in 2 5 miles NNE 11 5 mph 0 00 in
101. or module to store the data on a PC The second function of the XBee Pro module is to provide the logic for the voltage regulator circuit This is accomplished by utilizing the programmable sleeping capabilities The power supplied to the voltage regulator circuit is a 9V battery which supplies power to each component in the system Battery power has been utilized in order to minimize the need for AC outlets However if AC power is available the battery supply can be replaced with an AC power adapter I 1 Dylos ir Sensor V Battery RS 232 DB 9 Interface XBee Module 2 Dylos DC1100 Pro Laser Particle Counter The Dylos DC1100 Pro is an air quality sensor shown in Figure 2 contains two channels for capturing particle counts This device was mainly chosen due to its low cost but also for its dual channel design These channels have been calibrated by the Dylos Corporation to match the requirements of the system More specifically the calibration is important because it allows both PM2 5 and PM10 to be detected from a single sensor Figure 1 Interface Block Diagram of Sensor Node Device Details e Manufacturer Dylos Corporation dylosproducts com e Interface RS 232 DB 9 connector e Required Power 9V DC e Channel Calibration Channel 1 0 5 to 2 5um Channel 2 2 5 to 10 um e Cost 300 I 2 Figure 2 Dylos Air Sensor 3 XBee Pro DigiMesh 900 Module and Interface Board The XBee P
102. owNdx 1 Wend Close 1 Format Time Columns C C Select Selection NumberFormat 409 m d yy h mm AM PM Go to first cell Range A1 Select USK 2 2 EE I IC CE IC IC 2 KIC ICE aK KK I I ee aK 2 aK aK a Part 1b Import Location 1 Data txt to sheet Location 1 Data EEEE 21 1 EC C2 I EC a EI C1 A a EK IC I I ER 12 a CI A A a CI II a Ca a AR Ca ak aC a a ka Select sheet Sheets Location Data Select Clear current data in sheet Sheets Location Data Cells Clear Begin copying file on first cell Range A1 Select Application ScreenUpdating False On Error GoTo EndMacro SaveColINdx ActiveCell Column RowNdx ActiveCell Row File location in C Results FName C Documents and Settings amp UserName amp Desktop C Results Location Data txt C 28 Open FName For Input Access Read As 2 Sep w Use while loop to copy text file into worksheet While Not EOF 2 Line Input 2 WholeLine If Right WholeLine 1 lt gt Sep Then WholeLine WholeLine amp Sep End If ColNdx SaveColNdx Pos 1 NextPos InStr Pos WholeLine Sep While NextPos gt 1 TempVal Mid WholeLine Pos NextPos Pos Cells RowNdx ColNdx Value TempVal Pos NextPos 1 CoINdx CoINdx 1 NextPos InStr Pos WholeLine Sep Wend RowNdx RowNdx 1 Wend Close 2 Format Time Columns C C Select Selection NumberFormat 409 m d yy h mm AM PM Go to first cell Range A1
103. ration of Output File 1 Switch to the Capture Tab 2 In the file portion of the window type the destination of where the results text file is to be save o Current System C Documents and Settings Anthony Desktop RealTerm Output results txt 3 Make sure the Unix timestamp button is clicked so that a timestamp will appear in the final text file 4 To Capture the data click Start Overwrite As shown in Figure 16 Display Port Capture Pins Send Echo Pot 12c 120 2 12CMise Misc n Clear Freeze Capture aie Diagnostic Files Status Start Overwrite Start Append Bytes 0000000 7 Log hex _ Connected File C Program Files BEL Realterm results twt IS y gw RD 2 Fl Program Files ealterm results tx A lel ESE A _ TXD 3 Clear Dump M Capture as Hex TimeStamp Delimiter fies pre ICTS 6 C Matlab commal File E DCD 1 C YMDHS DSR 6 space realterm log El Ring 9 BREAK __ Error Doubleclick here to toggle Status information Figure 16 Realterm Capture Tab Details of Output File 1 The RealTerm output will be saved as a text file in specified location 2 The data will be in the following format e 1236366080 f H Ayp 1125 18 e An example of the file is shown in Figure 17 J 15 3 Explanation of output e 1236366080 Unix Timestamp which can be converted to readable time e H Node Identifier this will be different for each XBee module e 1
104. re Technical documentation has been included in this packet for reference The engineering manual is meant to direct personnel interested in reproducing the final system A user manual was written to instruct a potential client in the nuances of deploying the completed system in a live field 12 References 1 Dockery D W et al An Association Between Air Pollution and Mortality in Six US Cities New England Journal of Medicine vol 329 pp 1753 1759 1993 2 Samet J M Dominici F Curriero F C Coursac I Zeger S Fine Particulate Air and Mortality in 20 US Cities 1987 1994 New England Journal of Medicine vol 343 pp 1742 1749 2000 3 Gauderman W J et al 2004 The Effect of Air Pollution on Lung Development from 10 to 18 Years of Age New England Journal of Medicine vol 351 pp 1057 1067 2004 4 Pope II C A Burnett R T Thun M J Calle E E Krewski D Ito K Thurston G D Lung Cancer Cardiopulmonary Morality and Long term Exposure to Fine Particulate Air Pollution Journal of American Medical Association vol 287 pp 1132 1141 2002 5 Andrew Hass Environmental Protection Agency Private Communication 2009 6 Mark Winter Environmental Protection Agency Private Communication 2008 7 Andrew Hass Environmental Protection Agency Private Communication 2008 8 Paresh Mehta Air Management Service Private Communication 2009 9 Sales Represent
105. ro DigiMesh 900 is a newer series of XBee modules The Pro series features improvements in both range and power management Due to these characteristics as well as low cost the XBee Pro DigiMesh 900 module was chosen as the wireless component in the system In addition to the wireless module chip which contains an antenna two types of interface boards are needed in this system The base station which contains the PC for data collection requires a USB interface board The nodes each containing a Dylos sensor require a serial interface board All of these components can be obtained by purchasing an XBee Pro DigiMesh 900 development kit from Digi The circuit boards can be easily constructed by placing the XBee Pro module onto the receptacle of the interface board as shown in Figure 3 Alternative XBee kits may be purchased if they contain similar interface boards Device Details e Supplier Digi digi com e Interface Boards RS 232 DB 9 connector USB e Required Power 9V DC Serial Interface Board 1 3 Figure 3 XBee serial interface board with module circled 4 Null Modem Serial Cable RS 232 Interface In order to transmit the Dylos sensor output to the XBee Pro Module a serial communications link must be established This can be done by using the supplied null modem cable in the XBee Pro development kit Alternatively a null modem cable can be easily constructed with two DB 9 male connectors by following the pin out in
106. rough a USB interface board The coordinator is considered the data collector of the network and is directly connected to the Base Station or PC to allow proper data storage Installation Details 1 Double Click the X CTU software there should now be an icon on the desktop 2 Under the PC Settings Tab select the PC serial COM port that will be used The serial ports used will be labeled as Digi PKG U Serial Port Adapters 3 The settings should be as follows Baud Rate 9600 Flow Control None Data Bits 8 Parity None e Stop Bits 1 4 Once these are selected make sure the Enable API box is checked and click the Test Query button to determine if the correct COM Port was selected as shown in Figure 3 J 4 About PC Settings Range Test Terminal Modem Configuration m Com Port Setup Select Com Port Communications Port COM1 Baud 3600 X Communications Port COM Digi PKG U Serial Port Adap PAOA NONE Data Bits Parity Stop Bits C er Host Setup User Com Ports Network Interface AP V Enable API Use scape characters ATAP 2 AT command Setup ASCII Hex Command Character CC 7 2B Guard Time Before BT 1000 Guard Time After AT 1000 Modem Flash Update No baud change Figure 3 X CTU Com Port Setup Read Set Modem Parameters 1 Go to the Modem Configuration Tab 2 Verify the newest versions are instal
107. rs needed to improve the particle count to conversion algorithm Using this data the team performed an extensive analysis of various weather factors that could impact data this analysis was performed using a developed software package which gleaned historical weather data from WPHL17 television s website In addition to determining that the presence of rain cleaned the air and led to good correlation between collected data the team also determined that humidity Figure 14 Appendix D played a key role in mass underestimation These observations were supported by Lee et al who noted that in the presence of high humidity particles take on water and gain mass 20 This phenomenon would not be accounted for by a laser particle counter which simply counts the number of particles and thus needed to be factored into the particle count to concentration conversion algorithm The design team developed correction factors detailed in the next section to compensate for humidity and rain in its particle count to concentration conversion algorithm Figure 15 Appendix D shows the result of this modification a very strong correlation can be seen between the AMS continuous data and the design team s corrected data In an attempt to validate these modifications show that the algorithm is useful for a different sensor a second sensor was deployed at the AMS site Figure 16 Appendix D shows that the modifications are useful and acceptable on each
108. rs were utilized to augment this conversion method based on the presence of humidity and rain Specifically the improved algorithm can be described by the relation F O x H x C where O is the output of the initial conversion from particle count to concentration described above H is the relative humidity percentage C is the correction factor and F is the final output concentration Table 2 shows the values of H and C in dry no rain conditions while Table 3 shows the values of H and C in the presence of rain C 50 59 2 5 50 59 60 69 90 100 Table 2 H and C Factors in Dry Conditions Table 3 H and C Factors in Rain These factors were developed based on the comparison to the AMS continuous monitor and humidity data as described in the previous section The code needed to implement the conversion algorithm and to display data in appropriate fashion can be seen in Appendix C South Philadelphia Deployment The final step in the development of the preliminary screening system was deployment in the neighborhood specified by the Clean Air Council The goal of this deployment was to ensure that the system s main objective a short term low cost yet effective determination of the necessity of more comprehensive particulate matter testing could be achieved in a real environment Deployment was performed with the assistance of the Clean Air Council of Philadelphia who worked to obtain the permissi
109. rt laser particle counter data into data that can be compared against the NAAQS has been developed e Dylos DC1100 Pro laser particle counter was customized to evaluate the number of particles deemed hazardous those ranging from 0 5 2 5 um and 2 5 10 um by the Environmental Protection Agency e Experimentation performed using the Dylos DC1100 Pro yielded more precise power consumption data o Voltage regulator circuits to limit the amount of power consumed by the system and thus extend battery life have been designed and fabricated e Communication between nodes and a base station was achieved using Xbee wireless networking hardware o Data captured by the laser particle counter successfully transmitted to the base station e Appropriate documentation user and engineering manuals have been written The necessary system budget based on developments to date is 3 665 8 lower than the initial estimate of 4 000 Developments have proceeded on schedule and the system has been successfully deployed at the target location with the help of the Clean Air Council of Philadelphia In addition to the technical milestones achieved the design team had the opportunity to select and mentor a team of high school students from the Science Leadership Academy in Philadelphia PA as part of the Drexel Engineering Projects in Community Service EPICS Program The design team and high school students worked hand in hand through the engineering process which conclud
110. s DC1100 Pro Data gathered by the Dylos sensor is transmitted across a wireless network to a base station The wireless network utilized was the Digi XBee Pro DigiMesh 900 series Figure 3 selected for its minimal power consumption effective sleep mode and user friendly configuration settings Each node s sensor and wireless module is powered by an Impact BPD 8000 Rechargeable Battery Figure 5 Using the battery in conjunction with a voltage regulator circuit allows the node to operate in the field for approximately six days The base station utilized to collect data from each node is an Asus Eee PC Figure 4 The device features an on board monitor and keyboard minimizing the complexity of on site troubleshooting Loaded onto the base station is a developed software package used to appropriately analyze and display the collected data A data conversion algorithm was developed to convert from the Dylos sensor s output particle count 01 ft to concentration ug m as specified by the EPA standards The algorithm is coded into the software package more information can be found in the Particle Count to Concentration Conversion Algorithm section Throughout the design process specific deliverables have been followed toward system completion There have been several challenges which are also detailed in this report These challenges however have been addressed and resolved in the fabrication of the final system
111. s requested assistance in developing a system to be deployed in Southeast Philadelphia due to its proximity to Packer Marine Terminal the CAC has developed a relationship with neighborhoods within this region that have allowed for the deployment of the system 10 The work has also been performed with the help of a team of Science Leadership Academy high school students as part of the Drexel Engineering Projects in Community Service EPICS Program Final Design Due to the problems present in the current testing methods this system has been developed for preliminary screening of ambient air to determine the necessity of more comprehensive tests The system conserves costs time and manual labor and provides effective feedback for the selected neighborhoods in Southeast Philadelphia to determine if the EPA should be contacted for evaluation The system consists of both hardware and software components working together An overview of the system as well as a sample node can be seen in Figure 1 A more detailed technical analysis can be found in Appendix H The system s air quality sensor solution is the Dylos DC1100 Pro Figure 2 which collects a sample of air and determines how many particles within a specified size range are present in the air sample To compare to the NAAQS the two size ranges evaluated were 0 5 2 5 um and 2 5 10 um The Dylos sensor was an ideal solution due to its small size and low cost Figure 2 Dylo
112. st per Watt hour Voltage Regulator Circuit During both monitor and power off modes the air sensor pulled far too much current Thus a voltage regulator circuit was designed to act as a switch for the air sensor in order to minimize power losses in the system The voltage regulator acts as a switch turning on and off the current supplied to the air sensor This is done by taking advantage of the XBee s sleep mode The losses incurred from the voltage regulator are minimal to the system without it The voltage regulator circuit uses only about 8 mA in the conducting state and only about 5 uA in the non conducting state A schematic of the power for each node can be seen in Figure 10 Appendix H Node Placement Enclosures The nodes must be placed into a weatherproof enclosure to protect the hardware from damage There are many enclosures available but few that meet the specifications needed for this design The large volume occupied by the Dylos air sensor when placed face down in the enclosure resulted in the need for a very large weatherproof enclosure measuring approximately 7 inches deep 8 9 inches wide and 10 12 inches tall As a low cost solution the use of Primex P136 enclosures leftover from a previous system 11 was considered Though this could have lowered the operating budget the enclosures needed to be significantly altered to meet the specifications of the current system These alterations would have left the system q
113. stem Base Station Location sic scisect toasts iadccna agen diediedsieaaeaeaaiian E 1 Figure 21 Southeast Philadelphia Deployed Node Houston Community Center E 1 Figure 22 Senior Design Team Mentoring SLA Students ccc ecceesceeeeeeseeeteeeneeees F 1 Figure 23 Senior Design Team Performs a Sensor Height Test with SLA Students F 1 Figure 24 XBee Pro 900 Module Mounting on RS 232 Interface Board 06 H 1 Figure 25 Modem Configuration Screen for X CTU Software ccccccesceseeseeeteees H 1 Figure 26 Air Sensor Node Power Circuit Schematic cccccceecceeseeseeeeteceeeeeeeeeeeees H 2 iv List of Tables Table 1 EPA National Ambient Air Quality Standard 0 0 cec ce cecceeceeseeetseceteceteeeeeeeenees 2 Table 2 H and C Factors in Dry Conditions ccceeccceecceeseeeseeeeseceeeeeeeeeseecsaeceeenseeenaaes 6 Table 3 Hand C Factors in Ra Meaca oE E E E E 6 Table 4 Southeast Philadelphia Daily Particulate Matter Averages sessssesssesesseseeseese 8 Table 5 Comparison of Daily Averages Filter Method to AMS and ECE 19 Continuous Data Introduction and Problem Description Over the past two decades extensive research has been done to determine the deleterious health effects of fine particulate matter in ambient air on the human body These studies have shown various short term and long term effects resulting from exposure to P M jo particles less than 10 um in size and P M 25 particl
114. t SM SLEEP MODE to 4 Cyclic Sleep as shown in Figure 9 7 Click Write 8 The module will now be set to receive the sleep cycle from the coordinator 9 Repeat for all nodes in the network 17 X CTU COM3 Remote Configuration PC Settings Range Test Terminal Modem Configuration Modem Parameters and Firmware p Parameter View p Profile Versions Read Write Restore Clear Screen Save Download new Always update firmware Show Defaults Load versions Modem XBEE PRO Function Set Version MBPO9 DM v XBEE PRO DIGIMESH 900 x fisz0 B MO PWMO duty cycle f M1 PWM1 duty cycle B D PP RSSI PWM Timer D PR Pull up Resistor Enable Sy AT Command Options B CT AT Command Mode Timeout D GT Guard Times CC Command Sequence Character Sq Diagnostic Commands D VR Firmware Version B HY Hardware Version B 2 Supply Voltage D ER Receive Error Count D GD Receive Good Count D TR Delivery Failure Count 49 Sleep Commands D SO Sleep Options Bi 4 SM Sleep Mode v D ST Wake Period D SP Cyclic Sleep Period F f Normal mode is always awake Cyclic Sleep mode sleeps for SP time then wakes for ST time If D7 1 CTS flow control will follow sleep and wake periods COM3 9600 8 N 1 FLOW NONE Figure 9 Node Sleep Mode Configuration F Test Network Range A network range test should be done prior to deploying any hardware in the field This will help in determining your network cap
115. t endl Header Header lt lt setw 6 lt lt left lt lt Index lt lt setw 6 lt lt left lt lt WDay lt lt setw 17 lt lt left lt lt Date Time lt lt setw 6 lt lt left lt lt Node lt lt setw 36 lt lt left lt lt Node Location lt lt setw 13 lt lt left lt lt Humidity lt lt setw 12 lt lt left lt lt Conditions lt lt setw 13 lt lt left lt lt Small Count C 6 lt lt setw 13 lt lt left lt lt Large Count lt lt setw 14 lt lt left lt lt Conc PM 2 5 lt lt setw 13 lt lt left lt lt Conc PM 10 lt lt setw 8 lt lt left lt lt Message lt lt endl Location 4 ofstream outputLoc4 outputLoc4 open C Documents and Settings Anthony Desktop C Results Location 4 Data txt outputLoc4 Title lt lt setw 16 lt lt left lt lt Location 4 Data lt lt endl lt lt endl Header Header lt lt setw 6 lt lt left lt lt Index lt lt setw 6 lt lt left lt lt WDay lt lt setw 17 lt lt left lt lt Date Time lt lt setw 6 lt lt left lt lt Node lt lt setw 36 lt lt left lt lt Node Location lt lt setw 13 lt lt left lt lt Humidity lt lt setw 12 lt lt left lt lt Conditions lt lt setw 13 lt lt left lt lt Small Count lt lt setw 13 lt lt left lt lt Large Count lt lt setw 14 lt lt left lt lt
116. t setw 7 lt lt left lt lt Month lt lt setw 5 lt lt left lt lt Day lt lt setw 20 lt lt left lt lt PM 2 5 Average lt lt setw 20 lt lt left lt lt PM 10 Average lt lt setw 20 lt lt left lt lt Message lt lt endl Warning Report ofstream warning warning open C Documents and Settings Anthony Desktop C Results Warning Report txt warning Title lt lt setw 20 lt lt left lt lt Warning Report lt lt endl lt lt setw 20 lt lt left lt lt If PM 2 5 gt 35ug m 3 lt lt endl lt lt setw 20 lt lt left lt lt If PM 10 gt 150g m 3 lt lt endl lt lt endl Header lt lt setw 6 lt lt left lt lt WDay lt lt setw 17 lt lt left lt lt Date Time lt lt setw 36 lt lt left lt lt Node Location lt lt setw 13 lt lt left lt lt Humidity lt lt setw 12 lt lt left lt lt Conditions lt lt setw 20 lt lt left lt lt Conc PM 2 5 lt lt setw 19 lt lt left lt lt Conc PM 10 lt lt setw 9 lt lt left lt lt Message lt lt endl Error if input file does not exist if input fail cout lt lt File does not exist Please place file in correct folder lt lt endl cout lt lt Program will now end lt lt endl return 0 Data Point Index variables used to supply row number to each set of data int AllDataIndex 1 int Loc Index 1
117. t al Comparison of the ARB Continuous PM 5 Monitoring Network to the PM 5 Federal Reference Method Network California Environmental Protection Agency 2006 14 Appendix A Work Schedule Start Date End Date Time Duration Days al ie i3 P 2ilMatePesaaion S 2 2 Kickoff Meeting 3 9 2 3 SLA Team Presentation Meetings 1 30 Order Remaining Parts 1 5 3 9 1 30 2 14 3 13 12 a E F42jonierrars T 4 3 Inteface with solution Inteface with solution 2 9 Develop Voltage Regulator Circuit 2 9 2 27 3 6 asire Tore Gp nen 3a 12 F S Recach Ateri Parter S zi 52fFilze emit E EE a 34 17 EE io Figure 7 Winter Term Work Schedule Weeks 1 6 Month of January 2009 Month of February 2009 Week 1 Wek Wek Week U T Weeks PEE e Eee eee Ee Bee ae Week 2 A 1 Week 3 All Rob Falcone Anthony McClellan Ed Ostapowicz Mark Uva Week 6 Month of March 2009 Item Task Name Start Date End Date Time Duration Days 21 PF A WeeklyTeamMestings TT 2 1 Make Presentations 1 5 1 16 12 2 2 Kickoff Meeting 3 9 3 9 2 3 SLA Team Presentation Meetings 1 30 1 30 1 m e a ms 220 SE i EE 5 Interface With LPCs 2 16 2 27 12 4 Test Network 1 26 4 1 Research Power Options 4 2 Order Parts 4 3 Inteface with solution 4 4 Develop Volta
118. ta This is the first software that must run before analyzing the RealTerm output file results txt The Import Weather Data software is an important tool that automatically retrieves weather information from an internet source weather myphl17 com This is an Excel based tool that utilizes a VBA Macro User Input The user must first open the file on the desktop labeled Import Weather Data The user must choose the start and end dates as well as the closest location in order to capture the J 17 best weather information The user must make each choice through the use of pull down menus The user input page is shown in Figure 19 USER INPUTS Directions Please go through each step carefully before clicking the Import Weather button STEP 1 Choose the correct dates to obtain weather information The program will define invalid dates Such as 6 31 STEP 2 Choose the closest location to your network Currently only two locations exist NE Phila Airport and Philadelphia International Airport STEP 3 Once all information has been entered correctly click Import Weather to run program STEP 1 CHOOSE DATES STEP 2 CHOOSE STEP 3 RUN PROGRAM CLOSEST LOCATION Start Date Month Day Year Date Check Location VALID IMPORT WEATHER Testing Locations End Date Houston Community Center Philadelphia Int Airport Month Day Year Date Check Air Management Services AMS NE Philadelphia Airport VALID Figure 19 User Inputs f
119. thod filters and the EPA Federal Equivalent Method continuous particulate matter monitor previously discussed Because the filter data takes several weeks to obtain comparison to the EPA Federal Equivalent Method FEM was the primary factor used in improving the algorithm The continuous monitor utilized in the FEM comparison was the Met One Instruments BAM 1020 a large device which costs approximately 14 300 9 Studies have shown a very strong correlation between the FEM readings and the actual filter readings and concluded that both the FEM and the filter readings can be used for State regulatory purposes 21 The first sensor was deployed at the AMS site from April 3 2009 to May 5 2009 The second sensor was deployed at the site from April 21 2009 to May 5 2009 and was used to validate the improvements to the algorithm made using Sensor 1 Due to the two week lag time in the EPA s Federal Regulatory Method filter data the team was only given filter data from April 5 to April 16 2009 4 Figure 13 Appendix D compares the AMS data both the limited amount of filter data and the continuous monitor data to the design team s uncorrected data collected using Sensor 1 Uncorrected data refers to particle count data passed through the initially derived conversion algorithm described in the next section before the use of any correction factors This data would serve as the basis for the development of correction facto
120. tw 35 lt lt left lt lt location lt lt lt lt setw 12 lt lt left lt lt humidity lt lt lt lt setw 11 lt lt left lt lt condition lt lt lt lt setw 12 lt lt left lt lt smallCountInt lt lt lt lt setw 12 lt lt left lt lt largeCountInt lt lt lt lt setw 13 lt lt left lt lt conc25 lt an lt lt setw 12 lt lt left lt lt conc10 lt An lt lt setw 7 lt lt left lt lt message lt lt endl AllDataIndex Increase to next row value count data entries Warning Report string warningMessage if conc25 gt 35 amp amp conc10 gt 150 warningMessage Warning PM 2 5 and PM 10 too high Warn 1 PM2 5 over threshold elseif conc25 gt 35 warningMessage Warning PM 2 5 too high Warn 2 PM10 over threshold else if conc10 gt 150 warningMessage Warning PM 10 too high if conc25 gt 35 conc10 gt 150 warning lt lt setw 5 lt lt left lt lt weekDayID lt lt lt lt setw 16 lt lt left lt lt concatDateTime lt lt lt lt setw 35 lt lt left lt lt location lt lt lt lt setw 12 lt lt left lt lt humidity lt lt lt lt setw 11 lt lt left lt lt condition lt lt lt lt setw 13 lt lt left lt lt conc25 lt lt lt lt setw 12 lt lt left lt lt conc10 lt an lt lt setw 12 lt lt left lt lt warningMessage lt lt endl
121. udent Athlete 2004 e CYO Basketball All Star Selection 2004
122. uite vulnerable to weather hazards thus despite the possibility of cost savings an alternative solution was sought G 3 The option utilized in the final design was the Primex P1000 enclosure This product has a larger depth 4 27 inches which would allow the sensor to be entirely encased within the enclosure The enclosure also has sealable valves at its bottom which are opened to allow for airflow from the ambient environment into the box In addition to the larger depth this product also contains weatherproofing around the closing edges and is relatively inexpensive costing approximately 38 18 A plastic cylinder was placed over the mouth of the Dylos sensor to prevent the recirculation of air within the enclosure In addition to the node enclosures a separate enclosure is needed for the base station Since the base station does not include a Dylos sensor the depth is not a problem The chosen enclosure for the base stations was L Com s weather proof model NBP141004 100 This model has dimensions of 14 x 10 x 4 and costs approximately 45 As an added benefit a removable 120 VAC power module is included This allows easy access to power outlets for the Asus Eee PC 19 G 4 Appendix H Additional Figures X CTU COM Remote Configuration PC Settings Range Test Terminal Modem Configuration Modem Parameters and Firmware Profile Write Restore Clear Screen Save I Always update firmware Show Defaults Load
123. ul data This data will then be passed through a mass lt lt endl cout lt lt conversion algorithm All data will then be saved to a new file lt lt endl lt lt endl cout lt lt Location of RealTerm text file is in folder RealTerm Output lt lt endl cout lt lt Output format of new files in folder C Results lt lt endl lt lt endl Enter Set up Location 1 AMS or Demo Setup 2 Houston Community Center cout lt lt User Directions lt lt endl cout lt lt Enter 1 AMS lt lt endl cout lt lt Enter 2 Houston Community Center lt lt endl lt lt endl cout lt lt User enter number from choices above C 4 int choice cin gt gt choice cout lt lt endl lt lt endl EE EEEE EEEE EEE EE EEE EEEE EEEE E EEEE EEEE ISOS CISC ICICI IEICE Part 1 Read RealTerm output file results txt H Create New Folder and Files Data index variables defined for row numbering Count Variables used to justify daily averages PRB BRASS SCIFI EEEE EE CI CIS ICCC ACI 3 I EE E 3 A I I EI AE I I EC IC EE CA Read in RealTerm Output File ifstream input input open C Documents and Settings Anthony Desktop RealTerm Output results txt ios binary Create New Folder mkdir C Documents and Settings Anthony Desktop C Results New Output Files All Headers All Data ofstream outputAlIData File name outputAllData open C Documents a
124. wer cost than proposed Other cables included in XBee development kit Higher cost than proposed needed 2 one for AMS and one for BS Higher cost than proposed Included in XBee devlopment kit Not used in design ower cost than proposed ower cost than proposed Not used in design Needed for enclosure to secure parts Not in proposed budget Currently under budget Proposed Senior Design 2008 2009 ECE Team 19 Expense Report Actual Senior Design 2008 2009 ECE Team 19 Expense Report Category QTY Benefits Quarter 1 Quarter 2 Quarter 3 Category QTY Benefits Quarter 1 Quarter 2 Quarter 3 Total a a ST RT RN RN LR NY I GE Faculty Advisors 200 hr Faculty Advisors for 20 hours 10 400 00 10 400 00 i j 200 hr for 20 hours 30 10 400 00 10 400 00 10 400 00 Consultant Consultant 50 hr for 20 hours 1 300 00 1 300 00 50 hr for 20 hours 30 1 300 00 1 300 00 1 300 00 35 hr for 150 hours 4 30 27 300 00 27 300 00 27 300 00 81 900 00 35 hr for 150 hours 4 30 27 300 00 27 300 00 27 300 00 81 900 00 Freshman Engineers Freshman Engineers Hardware TT N a ce ES Laser Particle Dylos DC1100 Pro Counters 305 00 j Custom Calibration 304 94 Batteries y j Batteries Weatherproof Node Enclosures Enclosures j Base Station Enclosure Mounting Assemblies Wireless Transceivers Nodes and XBee Pro DigiMesh 900 Base Station 200 00 Antenna Celluar Connectio

Preliminary Screening System for Ambient Air

Contents

Download Pdf Manuals

Related Search

Related Contents