Home

of Figures

1. ad 16 35 HARDWARE 19 3 1 FORCE SENSORS ARRA 02 Rc LI EM 20 325 NIICROCONTEGEEER ea ut eeu A 27 3 3 CONDITIONING CIRCUIT nes 20 3 4 INERTIAL MEASUREMENT UNDE _ 31 3 5 COMMUNICATION aureis tius ode Gate vest 33 3 6 MICRO SD CARD MEMORY STORAGE sccsccscesceccecceccecescesceccscescescescecescescescesescescscescssescesceseeces 36 3 T ENTE CONS 2 2 cate pec DI CUT 37 3 8 PRINTED CIRCUIT BOARD 38 do 3 40 4 1 REQUIREMENTS timer en gate ra 41 4 2 LL AL uL LS eT RCE RE CE 41 4 3 IMPE EMENTA OR ores a 46 HARDWARE INTEGRATION TESTS oss isi rer NOR Yee URN
2. sensor 15 MD frente 5 2 d do l 23 Sensor sensor frente MD 2 2 16 MD frente 2 2 dado 1 26 5 Sensor irenteld 5 Sensor 8 5 2 MD frente 55 sensor 17 MD frente 4 3 1 21 sensor ire ente 4 2 sensor rencte 4 2 MD 1 3 sensor 18 MD frente 3 3 dado 1 14 7 110 sensor irente 4 4 sensor irente 4 4 MD frente 4 4 gsensor 19 5 MD frente 5 4 dado 1 20 sensor 3 2 sensor 3 MD rrente 3 2 37 Ssensor 20 MD frente 2 4 dado 1 11 neor rentel oyo MD Crentes al Sensor 21 MD dadoti l9 sensor frentelz 3 Sensor Irento s2 2 47 MD Erente 2 3 7 22 MD frente 5 5 dado 1 12 sensor trenteiZ d sensor 4 4 MD trente 2 4 7 sensor 23 MD Trente 5 6 dado l 22 sensor 134 sensor Irente tl 2 T MD Sssensor 24 MD rIrente 2 0 113 Sensor Frente 1 3 sensor frente l s 4 MD frentetd 5 data sensor interpolation NO tras interp MD ND interp MD 4 me ee ee 00122110 A SS spiren y
3. 68 9 2 PLANTAR PRESSURE 5 on d ac cde 72 5 3 SUREBOARD MOVEMENT ans iaa _ 75 CONCETUSION 5 5 2 ada E OIL d odd 77 6 1 EUTURE WORKS d eerie 79 REFERENCES 81 86 205 87 APPENDIX C CIRCUIT SCHEMATIC DIAGRAM cotta 88 APPENDIX D SENS 09268 SCHEMATIC DIAGRAM ccsscssssssssscsssscssccsssccssccssssscssscssssscsssees 94 APPENDIX E PRINTED CIRCUIT BOARD 95 APPENDIX F RIC FLOWCHAR E 98 APPENDIX G CALIBRATION EQUATIONS AND 1 1 1 1 1 100 APPENDIX DATA FRAME FIELD DESCRIPTION ccccssssssssssscccccccssssssscccccccsssssssccsssesees 106 APPENDIX I CENTER OF PRESSURE FIRMWARE CODBE ee eee eee eee ee eee eee eee es 107 APPENDIX Ji MATLAB CODE elio eee exero ev viv ia dcus da cad e cosa a 109 Index Figures Figure Take off movement Four different mome
4. 25 Table 3 Coefficient of correlation of 4 4 N force range sensors 25 Table 4 Measurement error verified after calibration 26 Table 5 Peripherals available in the AVR 1280 28 Table 6 Dynamic characteristics for HEF4OS 31 Table 7 Features from ADXL335 and DGS00 ais 33 Table S SPDcontectorns describe a bebe a a a etl hts 36 Table 9 Multiplexer pinout configuration 43 Table 10 UART 0 and UART 1 registers configuration 43 Table 11 Timers 1 and Timer 2 registers configuration ccccccccccceceecceeaaeeeesseeeeeeeeeeeeeeseeeeeaaas 44 Table 12 ADC registers configuration oio etae ede ieee 45 Table det coe oa ace aes foie 45 Table 14 SPI initialization pinout assigned 59 Table 15 CoP displacement and rotation along the simulation tests O 73 xiii Acronyms ASP GPS LE CoP IMU EEPROM SRAM RISC JTAG ADC UART SPI OPAMP PCB SMD USB LSB MSB Association of Surfing Professionals Global System Posit
5. 50 Figure 29 Timer 2 counter interrupt routine for RTC control 51 Figure 30 UART receive interrupt code routine 52 SU ART transmit eode EE cell ha aequ Ud voa UE cel vba 52 ADC store Tesi sequence ei ebd ict b oem e boo 53 Fiut Ss ADC IN 54 Figure 34 Multiplexer input select code 55 xi Figure 35 ADC Force sensors acquisition flowchart 56 Figure 36 Force sensor acquire extract function 56 Figure 37 ADC average calculation and off set 2 2 222 2 010 0005 5 57 Figure 38 ADC result to force conversion 58 Fiure conversion COde eter meteo oa ter uet tx cat ou aa 58 Figure 40 UART initialization parameters e Rn ae a A ao y ee deae auia 60 Figure 41 Transmit data through UART using FIFO buffer 60 Figure 42 Serial WiFi data read implementation 61 Data elds Seguent Poo od ebd ee d 61 Figure 44 Ac
6. Test 2 Pitch 7e Test 3 Roll 40 0 50 100 150 200 250 300 350 400 Samples Figure 56 Pitch and Roll along the Test 2 and Test 3 75 The test results have shown that both pitch and roll can be measured for the system at the same time Similarly during Test 1 it was observed the stabilization of pitch and roll around zero degree is reached It is also possible to verify the initial transients when the surfboard has its position changed by the surfer Figure 57 shows the pitch and roll during the first trial of Test 1 Surfboard Rotation on Test 1 20 4 Pu ciu 15 M 1 1 1 10 Test 1 Pitch 5 d Test 1 Roll 0 4 0 50 100 150 200 250 300 350 400 Samples Figure 57 Pitch and Roll along Test 1 76 6 CONCLUSION The use of technology is growing every day and in electronic field it is more visible due to the huge number of applications that are being developed either to make easier daily life or to help in medical diagnosis of diseases for instance The literature review has shown an existing gap in surf analysis although there have been efforts in order to minimize the gap between the practicing and practice evaluation Despite of been considered a relaxing activity surfing requires great physical effort from the surfer mainly due to the environmental conditions and the movements complexity Furthermore the surfer 15 highly susceptibl
7. zu 4 222 if usart 8 223 224 BufferUSART8 bufferTx BufferUSART8 endTxF data insert new data into buffer 225 226 if BufferUSART8 endTxF lt BUFTX LEN 2 update data buffer address 227 f 228 ButferUSART endTxF 229 238 else 231 232 BufferUSART8 endTxF 8 233 1 234 when an overflow occurs overwrite the oldest byte 235 if BufferUSART8 endTxF BufferUSARTB endTxI 236 237 if BufferUSART endTxI BUFRX LEN 2 238 239 ButferUSART endTxI 240 241 else 242 243 ButferUSART endTxI 6 244 245 1 245 247 UCSReB 1 lt lt UDRIE empty transmit buffer interrupt enable Figure 41 Transmit data through UART using FIFO buffer To process data received from receive interrupt two functions named as getSerialDataUSB and getSeralDataWiFi return the data stored in the FIFO RX buffer from UARTO and 60 respectively These functions verify if there is any remaining data to be read in buffer returning it after updating the buffer s addresses Figure 42 shows the code implemented 483 char getSerialDataWIFI void 484 405 char cData 8 466 487 if BufferUSARTI endRxF BufferUSARTl endRxI checks if there is data on buffer 488 cData BufferUSARTI bufferRx ButfferUSARTI endRxI read oldest data from buffer 469 410 if BufferUSARTI endRxI lt BUFTX 1 2 update data buffer address 411 BufferUSARTI1 endRxI Ht
8. 17 2 6 2 SURFBOARD ROTATION The surfboard can perform rotations about any axis in three dimensional 3D space during wave riding Furthermore it 1s recognized that these rotations can be performed independently from each other The surfboard s rotational axes are depicted in Figure 5 By convention for this project the axes x y and z are positioned longitudinally transversely and normal to the surfboard respectively The rotation s direction is positive when in the clockwise direction Figure 5 Surfboard s pitch roll rotation axis orientation In electronic systems developments it has been usual to use an electronic Inertial Measurement Unit IMU to measure rotation The IMU has at least one accelerometer and one gyroscope to provide information about acceleration and rotation about one axis however it is common to find IMU s that integrate other components e g barometer and compass Generally the signals from these devices are processed by a microcontroller or microprocessor to provide the information of rotation and direction In this project an IMU with 5 degrees of freedom measures accelerations on each direction X Y and Z and roll and pitch movements i e rotation along X and Y This data allows determining the surfboard s movement 18 3 HARDWARE The absence of studies characterizing surfing during its practice is noticeable The characteristics of the environment where is practiced a
9. Thr FECEOIE 11 131 000666316 08 5222 2222 LI 2 _ _ b model of the second version Figure 19 Printed circuit board 39 4 FIRMWARE Due to the interdependence between hardware and firmware both are concurrently developed however general tests and validation procedures only can be done when the development process is concluded Especially in firmware applications the features are fully hardware dependent due to the low level of the implementation Based on the V model development process the firmware development can be described by using five major steps requirements definition design code implementation hardware integration and validation as depicted in Figure 20 7 Requirements 7 Validation implementation Figure 20 Firmware development based on V Model process 1 V Model development process is methodology for software hardware development that demonstrates the relation between the phases of the development 40 Based on the V Model process presented the steps are described in the sequence 4 1 REQUIREMENTS DEFINITION The firmware is responsible for implementing a routine that controls the microcontroller operation in order to achieve the project goals and its requirements can be seen as tasks to do it Therefore th
10. gt 6 Delete File gt Select Option 0 6 E Figure 26 Debug menu options To update date and time functions updateTime and updateDate are called These functions read the actual date and time and update it according to the data typed by user There 15 also a type verification to certify that wrong characters are not inserted Those routines are implemented in routines file On the other hand to manage SD card files functions findFiles readFiles and deleteFiles are used from FAT32 file Once Acquisition mode 15 selected the system starts to run and process data from sensors To control the acquisition frequency a flag 15 used from Timer 1 When the flag signals a new acquisition functions that get data from sensors are called Those functions and flag will be explained later After that the SD card s file name 15 created and data converted 15 stored in dataString array in ASCII format Also the data is sent by WiFi using function sendWiFiData and stored in SD card using function writeFile 4 3 2 INTERRUPTION There are seven interrupt vectors implemented on this project firmware which controls ADC Timers 1 and 2 and UARTO and UARTI peripherals The ADC vector interrupt 1s executed always when an ADC conversion ends The interrupt routine reads the conversion result storage into two 8 bit registers ADCL and ADCH and converts them into a 16 bit integer variable named ADCResult The interru
11. 0 9994 Sensor 8 Resistance Voltage Measured v Secondly to ascertain the measurement error the second test was performed by loading the sensors with the same force range and checking the output conversion on a terminal The highest measurement error verified is 0 25 N for the range of 3 52 N on Sensor 18 which represents 7 18 on that range However the mean error is 0 012 0 064 N which is not significant in this application The Table 4 shows the mean measurement error for each sensor calculated from the response of the whole force range Table 4 Measurement error verified after calibration Sensor 3 4 5 6 7 9 9 10 11 12 Meas Error N 0 00732 0 03398 0 01843 0 03620 0 01398 0 02268 0 02491 0 01157 0 01935 0 02732 26 Sensor 15 16 17 18 19 20 21 22 23 24 Meas Error N 0 00287 0 00380 0 05843 0 08509 0 01732 0 00491 N A 0 01398 0 01935 0 01843 The Figure 12 shows the measurement error associated with the force applied for Sensor 18 and Sensor 3 Measurement Error 4 wr 3 5 4 lt gt eo 25 25 44 2 au Sensor 18 1 5 4 XC 3 Force Range 0 1 2 3 4 5 6 7 8 9 force step Figure 12 Measurement error associated with force applied 3 2 MICROCONTROLLER The microcontroller can be considered the main electronic component of the system since it s where proces
12. else 2 1 1 frame 1 1i else xr num2str xr total yr num2str yr total numzstr resultante total 01 Tau tts strcat a 2 Swrite data into a txt file Se ey Ali 4 0 else break stops when the total number of lines is achieved end end end 112
13. 2 ERA Figure 17 131 module with integrated on chip antenna 32 35 3 6 MICRO SD CARD MEMORY STORAGE To store the data a Micro SD card support was implemented using the Serial Peripheral Interface SPI peripheral available in the microcontroller The SPI allows full duplex communication between devices up to 4 MHz with 16 MHz main clock The Micro SD card 15 connected to the microcontroller using four control pins plus two power pins The controls pins are described in Table 8 Table 8 SPI connections describe Acronym Description Function MOSI Master Output Slave Input Send data from master to slave device MISO Master Input Slave Output Master receives data from slave device SCK SPI Clock Synchronization clock CS Chip Select Select the current chip to receive transmit data As hardware implemented peripheral the SPI protocol is controlled internally by the microcontroller and the peripherals registers and flags providing the information about the state of communication To control the micro SD card using SPI interface the micro SD card must be a slave and the microcontroller must be a master device That configuration allows the microcontroller routines control the read write process Due to these requirements MOSI SCK and CS pins are output pins and MISO 15 input pin from the microcontroller view The micro SD card socket connection 15 presented in Figure 18 D10
14. A Processing USB e Inputs Communication Figure 6 System block diagram 3 1 FORCE SENSORS ARRAY In order to measure the feet pressure it was necessary to define which technology would be used on this project To figure out the possibilities for this implementation a study on the available commercial force sensors was done of three well know distributors taking into consideration the application characteristics mounting surface the surfboard s deck and the available budget The search was restricted due to the following requirements e No Load Cells Although load cells could measure the force over three directions giving information about normal and shear force the construction of a force platform over the surfboard would be extremely tough work To do that the load cells should have one side fixed on the base of the platform and the other side fixed on mobile http www digikey com ou Digikey Corporation 2 http www mouser com ou Mouser Electronics http www farnel com ou Farnell 20 part of the platform and then to allow the foot positioning measurement it would be necessary to create an array of small platforms Furthermore the materials used in the production make the load cells heavy for this kind of application Overall the price of each load cell could be hundreds of dollars which would make the project not feasible No liquid Pressure Sensors Once upon starting
15. sqrt 1 squared Rw6yro squared Rw6yro 1 combine Accelerometer and gyro readings for w 8 w4 2 w RwEst w RwAcc w config woyro RwGyro w 1 contig w6yro Figure 45 Pitch and Roll estimation code MATLAB APPLICATION shows the graphical interface and pitch and roll angles during testing 64 MATLAB R2011b 52 Edit Text Go Cell Tools Debug Parallel Desktop Window Help E Figure1 1 File Edit View Insert Tools Desktop Window Help Od us BIS 7 4 2 08 af y axis cm og x axis cm New to MATLAB Watch this Video see Demos or read Getting Started x pitch roll 2 fx Busy script Ln 389 Col 32 OVR Figure 46 Matlab application viewing CoP and position information After opening a communication channel the data received is stored into a file in Matlab that is read afterwards and then separated and stored into local variables The variables that correspond to the force sensors are stored into two matrices which have the same distributions as that of the sensors in the surfboard These matrices are plotted in 2D charts The CoP data is plotted in 3D chart where X and Y axes represent the CoP position coordinate plane and Z axis represents the absolute force value The corresponding code is Shown in APPENDIX J 65 5 TESTS The tests performed aim to validate the system s hardware operation by anal
16. 412 else 413 BufferUSARTl endRxlI 414 415 416 return cData return data read 417 Figure 42 Serial WiFi data read implementation code Due to a demand for a standard data frame transmission the function sendData creates a frame that is composed by as a frame start character followed by data stored in dataString variable and a command The frame sequence 15 shown in Figure 43 and field description is show in APPENDIX H Figure 43 Data frame fields sequence 4 3 6 CENTER OF PRESSURE The center of pressure calculation 15 based on the data from force sensors and the position of these sensors on the matrices Yr Both matrices have the sensors equally spaced and the distance between them are fixed The equation to calculate the CoP considers an individual force contribution F and distance from each sensor X Y over the total resultant force Take into consideration that the vector of force s direction 15 61 normal to the matrix plane Equation 5 provides the resultant force vector position Yr on each platform Fp X EUN 5 R Yp LCF x However to determine the body center of pressure a new calculation must be done considering those individual contributions Equation 6 uses the values of Xp Yp and Fp from rear and front platforms Frrront Frrear Frrront X Xnrront Frrear X Xnnear XTota
17. 7 22 Table 7 Features from ADXL335 and IDG500 ADXL335 Parameter Condition Typical Unit Input Signal Each axis Operational range 3 6 g Sensitivity Each axis Xout Yout Zout Vs 3V 300 mV g Voltage in Og Xout Yout Vs 1 5 1 65 V Zout Vs 3V 1 5 1 8 Max Supply 1 8 3 6 Max IDG500 Operational range X OUT e Y OUT 500 9 5 Sensitivity X OUT e Y OUT 2 0 mV 9 s Voltage reference 1 35 V Supply 2 7 3 6 Max V 3 5 COMMUNICATION WIFI AND USB To allow the system to communicate with the user two different ways have been provided WiFi and USB interfaces First of all USB interface 1s used as a debug interface Second WiFi transfers data acquisition and receives data commands over the air allowing the application to work remotely Both ways were implemented using external converters connected to the microcontroller s UART channels one FTDI232R 31 and one RN 131 32 modules communicating with the microcontroller through and UARTI for USB and WiFi connections respectively Once the modules were connected the communication protocols were implemented by the modules being transparent from the microcontroller s programmer view The connections between USB converter FTDI232R and microcontroller are shown in Figure 15 Due to the fact of not implementing the flux control a few connections had to be done to provide a UART USB conversion in which the signals labeled as FTD R
18. Figure 18 Micro SD socket connections 36 3 7 TIME CONSTRAINTS Due to the tasks performed during the conversion and processing data the time constraints is the result of the peripheral s clock frequency and or the code program implemented The analog to digital conversion routine converts 24 force sensors Nrorce plus 5 IMU Nimu signals Furthermore force sensors conversion routine uses an average filter AVG to smooth the signal filtering high oscillations Due to the ADC characteristics and clock frequency the minimum period for conversion 15 104us given by Equation 2 T 13 x 2 Conv ADC E Thus the total conversion time 15 21108 considering times from force sensors IMU and filtering process given by Equation 3 Trorce Tcony X AVG X Neorce Timu T conv X Nimu 3 TotalADC T Force Due to an UART baud rate BRateyarr of 115200 bps and the total number of sent bytes Npytes of 231 see section 4 3 5 Data Communication the UART time constraint is approximately 16 ms given by Equation 4 11 TR Due to successive approximation circuitry a normal conversion requires 13 clock cycles to be done The ADC clock frequency is defined during peripheral initialization process for 125 KHz 37 UART x 8 x 4 BRateyaprr Apart from those constraints the highest time constraint i
19. PORTA 1 SEL C else PORTA amp 1 lt lt 5 1 delay us 18 delay 71 72 if 1 lt 7 73 74 set selABC 5 input Y7 75 PORTD 1 SEL A OUT 76 PORTD 1 lt lt SEL B OUT 77 PORTD 1 lt lt SEL_C OUT 78 79 delay to stabilize values on delay us 1 delay 61 82 disable mux output active low 83 PORTA 1 EN MUX 2 84 PORTA 1 lt lt MUX 3 85 delay_us 1 delay 66 87 enable output active low 88 PORTA 8 1 lt lt 1 enable mux 1 89 delay us 1 delay 91 enable 5th output active low 92 PORTA amp 1 lt lt EN MUX 5 93 94 delay to stabilize signals 95 delay_us 1 delay 96 Figure 34 Multiplexer input select code extract 4 3 3 3 Acquiring sensor s signal To acquire data from force sensors and IMU two different functions are implemented where getForceSensors gets data from force sensors and getAccelGyro gets data from IMU The first one selects the ADC channel 14 converts the value implements an average filter and converts the result into Newton meanwhile the second one sweeping between the ADC channels 9 to 13 converts the value and store it Despite of being used in different manner both functions present the same structure for ADC conversion and only the force sensors conversion will be explained in detail The code flowchart is shown in F
20. SD card initialization messages 47 4 3 1 2 OPERATION MODE SELECTION Afterwards there are two modes that control system operation which are Debug and Acquisition modes Using a menu that is accessed through a communication terminal the user selects in which mode the system is going to operate by selecting option 1 to Debug mode and option 2 to Acquisition mode as depicted in Figure 25 If an invalid option 15 selected the user receives an invalid option message and the menu 15 refreshed enabling new selection 2 COM amp 115200baud Tera Term VT File Edit Setup Control Window Help Standard Capacity Card 2 x Detected Press kev Select the operation mode gt 1 Debug mode gt 2 Acquisition mode gt Select Option 8 2 E Figure 25 Mode menu selection Once Debug mode is selected a new menu shows the mode options which the user can select to configure system parameters Figure 26 The option zero exits from the menu and back to the menu mode selection Option 1 shows the current date and time meanwhile options 2 and 3 update its values Furthermore options 4 to 6 manage SD card file 46 COM8 115200baud Tera Term VT File Edit Setup Control Window Help gt Select Option 8 2 1 Exit the Menu gt 1 Display current Date Time gt 2 Update Date gt 3 Update Time gt Get file list gt 9 Read File
21. Technologies Figure 48 Hardware test setup 5 1 1 TIMING The timing verification was divided into two tests in which ADC acquisition frequency and multiplexer time constraints First of all ADC acquisition frequency was verified by acquiring the enable signal from the multiplexer due to this signal being only activated when new data acquisition routine takes place The multiplexer enable signal 15 active low Figure 49 shows the oscilloscope screenshot with two enable signals showing also the mean verified frequency of 10 052 Hz 68 High Res 900kSa s 202213 T ss Dev Freg 1 1 21 E E 2 7 362Hz 10 930 2 96mHz 32 10 516uz E HIT Low 27 ee Figure 49 ADC frequency verification Thereafter the tests for enable timing verification were performed by analyzing the time between enable disable enable instants The enable disable analysis was performed acquiring data from multiplexers and 2 during transition between multiplexer 1 enable and multiplexer 2 disable The results have shown an average time interval of 9 935 us which is depicted in Figure 50 The channel 1 on the oscilloscope represents the multiplexer 1 and channel 2 represents the multiplexer 2 Due to the duality between enable and disable time routines the results for disable enable are similar 0 On tests context is considere
22. distribution With the array of force sensors measuring the surfboard s reaction force on each foot it 1s possible to determine the feet s positions On the other hand to measure surfboard s angle it is necessary to use a different kind of sensors that provide information about acceleration and rotation converted into angles and 3 The force platforms are composed by hard materials like aluminum and steel measure the force on three directions in which normal force and shear forces Furthermore the force platforms are commonly installed over hard support to minimize noise from undesired vibrations from the surrounding structure 16 then the surfboard s movement can be determined Figure 3 shows the system components and their orientation rear platform electronic front platform system Figure 3 System positioning over the surfboard and surfboard s orientation 2 6 1 CENTER OF PRESSURE To measure the plantar pressure force sensors based mainly on capacitive and resistive technologies have been used The sensing elements are normally arranged to form a matrix of sensors as shown in Figure 4 from Tekscan Incorporated USA that uses force sensing resistors FSR as the sensing element Figure 4 Matrix of force sensor elements to measure pressure distribution by Tekscan Depending on the matrix space resolution it is possible to have enough accuracy to define feet position precisely
23. features to be considered on the multiplexer circuit are the serial equivalent resistance and propagation delays When an input 15 selected the sensor signal must pass over two multiplexers to be connected to the OPAMP circuit associating multiplexer resistances to the sensor s resistance Although the minimum supply voltage accepted 15 3VDC the component datasheet do not specify a typical resistance curve for that range Take into consideration the typical curve presented for 5VDC the resistance on conduction is typically 115 350 Q In worst case total resistance associated by two multiplexers will be 700 Q Due to the sensor resistance specified by the manufacture being larger than 5 when there is no load applied and the lowest sensor resistance measured of 24 8 in maximum load the associated conduction resistance can be ignored http www nxp com documents data_sheet HEF4051B pdf 30 On the other hand the propagation delays must be considered during firmware development to ensure proper analog to digital conversion However similarly in the conduction resistance the propagation delays are based on 5VDC supply voltage Table 6 shows the delays where HIGH to LOW and LOW to HIGH are the delay for transitions between high and low voltage levels HIGH to OFF OFF to HIGH and LOW to OFF OFF to LOW are the time for signal stabilization after switching Table 6 Dynamic characteristics for HEF405 Parameter Sig
24. for all axes and the necessary connection to acquire the electrical signals As a prototyping kit the signals are accessible for standard 2 54mm pitch block terminal connectors which makes easy to incorporate into a printed circuit board PCB However there are dimensions requirements that must be considered during the PCB design The SENS 09268 is shown in Figure 14 and the electrical circuit is on APPENDIX D ADXL335 Figure 14 Inertial Measurement Unit SENS 09268 from Skparkfun Electronics The electrical signals provided by the IMU are analog signals in which the voltage varies according to the device movement For the accelerometer the signals represent acceleration over each axis Once on Earth all bodies are subject to the gravity force G force which represents an acceleration of 9 8 m s and must be considered on the accelerometer device measurements On the other hand the gyroscope signals represent a rotation around the axis and they are not affected by the G force When a rotation about any axis 1s performed the resultant Coriolis effect causes a vibration that is detected by the gyroscope system and then the converted signal 1s proportional to the rotation Both devices are connected to the circuit through the analog to digital converter from the microcontroller The signals are then read and the conversions are based on features presented on the datasheets The most important features are described in Table
25. input 3 3 CONDITIONING CIRCUIT The conditioning circuit 15 responsible for sweeping receiving and adjusting the signals from the sensors platforms to be acquired before the ADC The circuit is composed by an analog multiplexer cascade and an operational amplifier OPAMP Due to the need of using a matrix with 32 sensors five 8 1 multiplexers were cascaded where four of them UI to U4 represent the inputs and the last U5 combines them for one single output Afterwards the multiplexed output is connected to non inverting OPAMP Figure 13 shows one multiplexer and OPAMP circuit from the conditioning circuit 29 R16 1K5 196 Vsys C8 10nF U4 SEL A 11 16 SEL B 10 B SEL_C 9 100 13 Sensor_25 MUX 4 6 INH 10 1 14 Sensor 26 OUT MUX 4 3 10 COM 102 15 Sensor_27 103 12 Sensor 28 R15 104 Sensor 29 x OUT SENS 3 Sensor 30 i 105 55 E GND 106 E LM358DR GND 7 ee HCF4051BMI Figure 13 Conditioning circuit implementation The multiplexer circuit uses a HEF4051B 8 channel multiplexer demultiplexer components This component is actually bidirectional however was configured as a multiplexer with eight inputs IOO to IO7 and an output IOCOM Furthermore there are three signals to control the input addressing A B and and one active low signal to enable disable the multiplexer INH Due to the application in cascade topology the main
26. into different files grouping the function according to the functionality Next will be described the main code programming contents into those files 4 3 1 MAIN The main function controls the system setup and operational modes First of all a setup function is called to configure peripherals and initialize variables The peripherals are initialized in this function to ensure that all of them will start working before the main code starts to run then Timers UARTs ADC SPI and I O pins are initialized according to system requirements The multiplexer and IMU filter variables are also initialized Furthermore a force sensors calibration is performed 4 3 1 1 MICRO SD CARD INITIALIZATION After setup an infinite loop contains the code to run the main routine Due to the use of a SD card as a storage memory the function starts initializing it providing initialization messages to user and if the SD card is detected and the initialization had been successful the program will still be running Otherwise if some problem is detected the user receives an error message and the system will stop running The SD card initialization procedure is shown in Figure 24 2 COMB 115200baud Tera Term VT o mm 2 COMB 115200baud Tera Term VT 2 Ex File Edit Setup Control Window Help File Edit Setup Control Window Help SD card not detected Standard Capacity Card Ver 2 x Detected Press any key Figure 24
27. of Sports Sciences 2010 28 7 p 797 803 14 Mendez Villanueva A et al Inaccuracy of the HR Reserve vs VO2 Reserve Relationship during Prone Arm paddling Exercise in Surfboard Riders Journal of Physiological Anthropology 2010 29 6 p 189 195 15 R and S G d Santos Judging criteria in international professional surfing championships Revista Brasileira de Cineantropometria amp Desempenho Humano 2012 14 4 p 439 449 16 Mendez Villanueva A D Bishop and P Hamer Activity Profile of World Class Professional Surfers During Competition Acase Study The Journal of Strength amp Conditioning Research 2006 20 3 p 477 482 17 Professionals Association of Surfing cited 2013 10 11 Available from http www aspworldtour com ASP rule book pdf gt 2013 18 Mendez Villanueva A and D Bishop Physiological aspects of surfboard riding performance Sports Medicine 2005 35 1 p 55 70 191 Meir R A Lowdon and Davie Heart rates and estimated energy expenditure during recreational surfing Aust J Sci Med Sport 1991 23 p 70 4 82 20 Lowdon B J et al Manoeuvres Used and Judges Scores in an International Surfing Contest Summary Report 1996 National Sports Research Centre 22 Mendez Villanueva A Mujika and D Bishop Variability of competitive performance assessment of elite surfboard riders The Journal of Strength amp Conditionin
28. pitch dado 27 5 Sroll eixo x roll dado 28 Sacelerac o eixo x 1 dado 29 m s2 1 no eixo y accely dado 30 Im sz 1 no eixo 7 accelZ dado 31 m s2 eee reus 5 xr total dado 52 yr total resultante total dado 54 Centro de Press o Pl xr total yr total 0 Pa xr total yr total resultante totall pts 1 P2 T D LC E 5 Plot 3D for CoP if 2165 CoP Figure 1 Subplor l 3 1 3 ploto qptsit 2 DLS PS io MarkerFaceColor b Linewidth 2 axis 0 19 18 90 0 60 view 54 32 xlabel x axis cm vlabeli yeax35 6 1 7 N grid end splot 2D for each platform it plot decks LOU ly 2 213 pcolor MD tras Stitle rear foot caxis O 151 5 shading interp SUDDIOL L3p3 3 111 peolor MD frente Stitle front foot 10 151 5 shading interp Sdrawnow end splot 2D KY for position 1 plot posic o subplot 1 5 4 Gx Gx roll plot Gx 1801 subplot 1 5 95 Gy Gy pitch 1200 180 title end end Ssave frame data into a local file n size frame for i 1 2 1 if i lt n 2 1 if frame l 1 2
29. sequence defined The multiplexer circuit selects one force sensor at a time which is connected to the ADC through the channel 14 However there are five remaining signals that are connected to the ADC which represents the IMU signals from channels 9 to 13 The result of the force sensors conversion is a double type variable meanwhile IMU result is an unsigned integer type The variables are named dSensorValue N_SENSORS dResultantFrontX dResultantFrontY dResultantFront e dResultantRearX dResultantRearY dResultantRear e dResultantTotalX dResultantTotalY dResultantTotal e uiAccelXYZGiroXY N_ACCEL where N SENSORS and N_ACCEL the number of 24 force sensors and 5 IMU signals respectively 53 4 3 3 1 Selecting ADC channel To select the ADC channel a function named ADCSelectChannel receives the variable uiChannel as parameter Then the variable is used to assign the value into the register ADMUX Since the channels ADC9 to ADC14 only those 3 bits have to be changed Figure 33 shows the code to select ADC channel 156 Elvoid ADCSelectChannel uint8 t uiChannel 157 4 158 ADC input channel config mux5 8 188118 159 uiChannel amp 68x807 168 uint8 t uiChannelLow uiChannel amp 6x87 161 162 char a uiChannel amp 1 163 char b uiChannel amp 2 164 char c uiChannel amp 4 155 166 Jichar channelHigh uiChannel amp x2 gt gt 5 167 ADCSRB 1 MUX
30. the searching for force sensors it 15 unavoidable not to look for pressure sensors which measure the force applied by a fluid over a known area that gives the information about the fluid pressure Although a lot of options of pressure sensors appeared during the search the use of a fluid to measure the force in this project requires a huge effort to create a matrix of dots where each one has a fluid passing through it and then the fluid pressure have to be measured by the sensors Furthermore the number of pressure sensors must be the same of the dots on the matrix which also would make the project financially inappropriate e Individual sensors to form a matrix As previously defined it is necessary to implement a matrix of sensors to measure the feet position Although there are some companies that produce that sort of matrices the price 15 high Furthermore the companies typically do not sell only the matrices they sell an entire solution which includes hardware and software To reduce the cost of the electronics for the prototype individual sensors to implement a matrix was the chosen solution These requirements have limited the research s result remaining a few options including Force Sensing Resistor FSR and Flexiforce Sensor Both sensors could be applied on this project however the Flexiforce was chosen Two force ranges was selected for comparison in order to test which one best suit for the project 4 4 and 110 N The
31. the system validation tests and results Finally Chapter 7 presents the conclusions and identifies some options for future work 2 SURF CHARACTERIZATION 2 1 INTRODUCTION During the last decade surfboard ridding has seen an increasing number of followers and suffered an increasing attention from the media This fact takes the sport to another level attracting much more investment for this sector mainly in surf wear fashion which was condemned on the past because of the prejudice about surfers and their lifestyle Different from most Olympic sports although its practice 15 equally older surfing has a recent history as a professional sport dating from the 80 s the beginning of the professional contests 1 Apart from that the sport of riding waves also follows the technology evolution Different kinds of materials have been applied in different situations in order to help surfers to increase performance Furthermore the use of technology during sports practice minimizes the difference between the coach and athlete s perceptions on the athlete s performance 2 since by comparing the expected pattern with the movement performed both athlete and coach are learning 3 Due to that surfing has been following the trend of using technology to increase the surfer s performance although the major commercial developments focus on improving equipment quality like surfboards and wetsuits for example One of those improveme
32. to thank my wife Cynthia de Mello de Bona for the support and relief in hard times I would like a special thanks to Maria Arcelina Marques and Miguel Velhote Correia my supervisors for believing in this project and for the many hours spent in conversations about it I also would like to thank Gustavo Ribeiro Alves PhD to the referrals in the beginning To the institutions that somehow supported me during the development School of Engineering Polytechnic of Porto Biomechanics Laboratory Porto University Faculty of Engineering Porto University SRS Surfboards To my friends that due to the distance I left behind and for the friends that I have made in Portugal Especially those for whom this project was a common commitment namely Marcio Borgonovo dos Santos and Marcelo Peduzzi de Castro Finally but not least I would like to thank my co workers from Federal Institute of Education Science and Technology of Santa Catarina to trust in my work and support my absence during that time Resumo uso da tecnologia tem crescido nas ltimas d cadas nas mais diversas reas seja na industria ou no dia a dia cada vez mais evidente os beneficios que traz No desporto nao diferente Cada dia surgem novos desenvolvimentos objetivando melhoria do desempenho dos praticantes de atividades fisicas possibilitando atingir resultados nunca antes pensados Al m disto a utiliza o da tecnologia no desporto permite a o
33. turn on the wave requires a dorsal flexion of the feet rotation of hip and shoulders to the left On the other hand a right turn with a regular foot position involves plantar flexion and rotation of the hip and shoulders to the right Another factor that must be observed is the acceleration because controlling the acceleration the surfer is able to ride the wave and execute different maneuvers among other elements listed In order to increase de acceleration the surfer can make fine turn up and down the wave using body movements or can move the surfboard with flexion extension movements using the hip and knee Moreover if the surfer would like to reduce the velocity he can put more weight on the surfboard s tail or make strong curves It is also important to observe that the surfer can ride a wave in different directions in relation of the wave s face Basically there are two directions left and right Left wave occurs when the wave break from the left to the right in relation a one observer on the beach On the other hand right waves occur when the wave breaks from right to left Peirao et al analyzed two professional international championship events with 21 competitors from different countries by studying variables included in the ASP judging criteria elements and major manoeuvers the surfer position in relation to the wave length of the ride and frequency of major manoeuvers The major manoeuvers were described as Carving Re e
34. 3 1 35 6 682 133 dRyGyro double uiAccelXYzGiroXY GYROY AXIS 3 3 1823 1 35 0 002 134 E 143 normalize vector convert to a vector with same direction and with length 1 144 normalize3DVector RwAcc 145 Figure 44 Acceleration and gyro unit conversion The first estimation takes only the acceleration value as a reference and afterwards every interaction uses the data from IMU When previous estimated z angle is less than 0 1 it means that there is no significant variation and the new gyroscope data 15 loaded with the previous estimated value Otherwise the estimation algorithm gets the angle projection on plane ZX and ZY based on last estimation values and then updates the angle using actual gyro data Finally the accelerometer acquired data and gyroscope calculated data are used to estimate pitch and roll Figure 45 shows the principal code in the file to get angle estimation The sensitivity for accelerometer 15 330mV g meanwhile for gyroscope is 2mV s 63 145 149 158 151 152 153 154 155 156 157 155 159 168 161 162 163 164 165 166 167 168 169 178 171 172 173 174 175 176 177 178 179 150 191 182 183 184 185 186 187 185 4 3 8 A Matlab application was developed to allow real time analysis and visualization during test procedure The application opens a communication channel which allows reading the data frame and processing the data to show CoP and forc
35. 36 GND GND GND GND GND GND GND GND GND SENSOR 0 SENSE PWR SENSOR 3 SENSOR 2 SENSOR 1 GP GP GP GP GP GP DMA 3 3V REG OUT SPI MISO SPI CLK SPI MOSI RESET EPC ANT A EPC ANT B SUPERCAP GPIO 13 GPIO 12 UART RX UART TX RN 131 ladi _FORCE AWAKE SENSOR 6 8 5 4 3 SENSOR 5 a SENSOR 7 x 9 LO AA STATUS RESET ANTEN_A WIFILRX WIFI TX 2 cel R3 Figure 16 RN 131 schematic connection The maximum transfer data rate over the WiFi connection for this module is 54 Mbps however the UART connection presents an lower rate The maximum baud rate described on the component datasheet was Mbps for UART connection On the other hand the user manual presents different values were the maximum baud rate is 921 600 bps That sort of value cannot be configured in the microcontroller been impossible to reach the fastest baud rate on both devices Due to that the maximum rate reached was 115200 bps remaining implemented To connect the module wireless an on chip antenna is integrated on the WiFi module However it 15 possible to configure the module to use external antenna through the U FL connector allowing optimize system range The on chip antenna presents a gain of 2 0 dBi with maximum peak power of 237 7mW The module is shown in Figure 17 on
36. 5 168 169 if a ADMUX 1 lt lt else ADMUX amp 1 178 if b ADMUX 1 MUX1 else ADMUX amp 1 MUX1 171 if c ADMUX 1 MUX2 else ADMUX amp 1 MUX2 172 173 ADMUX uiChannelLow 174 Figure 33 ADC channel selection 4 3 3 2 Selecting multiplexer channel To select the multiplexer channel a function selectMuxChannel receives two parameters First parameter is the channel number and second is a command to enable or disable the corresponding channel Due to the 24 inputs the sweeping sequence uses three selection signals which are SEL SEL and SEL C for the first part and SEL OUT SEL B OUT and SEL C OUT for the second part of the cascade circuit By controlling the time constraints the channel selection and multiplexer enable signals are synchronized to ensure proper signal stabilization Figure 34 shows a part of the code that selects the first eight channels from the multiplexers The complete input sequence selection to assign correctly ADC channel using bits MUXO to MUXS can be consulted on component datasheet on table 26 4 16 Due to a binary system the number of bits to select 8 inputs is given by 2 where n is the number of bits For this case 2 8 54 65 66 Set selABC value 67 if a PORTA 1 lt lt SEL_A else PORTA amp e 1 SEL A 68 if b PORTA 1 sEL B else PORTA amp e 1 SEL B 69 if c
37. 6 Po 1 ll MON MEET NYC TU NY CN 7 ENT T Gamm 96 APPENDIX ISR_RTC Update local variables Update in progress yes RTC flowchart Miliseconds 10002 Miliseconds 1 Second 1 Miliseconds 0 Second 60 Minute 1 Seconds 0 Minute 60 66 Day 31 and month 4 6 9 11 UpDate 1 5 Check Leap Year Month 2 and Leap Year UpDate 1 UpDate 1 Month 1 Day 1 Month 13 Year 1 Month 1 001 APPENDIX G Calibration equations and charts Following are presented the data used to on each sensor s calibration The charts represent Force versus Voltage data from the sensor measured on the output of the conditioning circuit likewise the coefficient of correlation R and the calibration equation As previously described it is noticeable that the Sensor 21 presented the lowest R due to the shape of the curve Sensor 3 Sensor 4 45 Sensor x Sensor4 Linear Sensor 4 Linear Sensor 3 y 2 8776x 0 5538 3 1349 0 6907 R 0 9952 R 0 9903 01 Sensor 5 x Sensor 5 Linear Sensor 5 Sensor 7 Sensor 7 Linear Sensor 7 y 2 4783x 0 1044 R 0 9888 y 2 9021x 0 3926 R 0 9893 Sensor 6 x Sensor6 Linear Sensor 6 Sensor 8 Sensor 8 Linear Sensor 8 y 2 5768x 0 2185 R 0 9969 y 2
38. 6214x 0 3475 R 0 9994 Sensor 9 Sensor 9 Linear Sensor 9 Sensor 11 Sensor 11 Linear Sensor 11 2 0471x 0 1571 0 9899 y 5 2124 0 4655 R 0 9987 Sensor 10 Sensor 10 Linear Sensor 10 Sensor 12 Sensor 12 Linear Sensor 12 y 2 0796x 0 2489 R 0 9962 y 3 9748 0 5017 R 0 9919 Sensor 15 X Sensor 15 Linear Sensor 15 Sensor 17 Sensor 17 Linear Sensor 17 y 2 3896x 0 3593 R 0 9948 y 2 543x 0 5742 R 0 9862 Sensor 16 X Sensor 16 Linear Sensor 16 Sensor 18 X Sensor 18 Linear Sensor 18 y 2 5027x 0 3948 R 0 9929 y 3 7059x 0 3469 R 0 9976 Sensor 19 Sensor 19 Linear Sensor 19 Sensor 21 X Sensor 21 Linear Sensor 21 y 2 5377x 0 5839 R 0 9971 y 2 3053x 0 7867 R 0 9393 Sensor 20 Sensor 20 Linear Sensor 20 Sensor 22 X Sensor 22 Linear Sensor 22 y 2 7759x 0 5734 R 0 9859 y 2 55x 0 1121 R 0 9989 501 Sensor 23 Sensor 23 Linear Sensor 23 3 2409x 0 2353 R 0 9918 Sensor 24 x Sensor 24 Linear Sensor 24 y 3 593x 0 0286 R 0 9888 APPENDIX H Data frame field description The frame is composed by 34 fields where each one is described below Field 55 24 SENSORS PITCH ROLL ACCELX ACCELY ACCELZ COPX COPY COP
39. 68 from Skparkfun Electronics 32 1 gt PPDIZ3ZR Schematic CONDE CON 34 Figure schematic connections GT 35 Figure 17 RN 131 module with integrated on chip antenna 33 35 Figure 18 Micro SD socket connections Getae 36 Figure Printed creuit hoard EE 39 Figure 20 Firmware development based on V Model 5 22220000000001 40 21 gt General Mrmw are a SO 41 Figure 22 Setup sequence for system initialization 42 Figure 23 Modes of operation from main 46 Heure 24 SD cardanitralization 116858 65 5 2 2 0 47 Fieure 2 Mode Menu 8 00 08 48 Fioure 2060 D bus nien OpDOHS REIR E IS T cien 49 Figure 27 ADC read conversion registers sequence 00000000 nennen nnns 50 29 limer interrupt FOULITIB
40. 71 and it is given by Equation 1 where r represents sensor s radius of 0 4765 cm 22 Asensor X p 1 Due to the sensor s range of 4 4N to LION and area calculated the maximum measureable pressure is 6 17 and 154 21 respectively In what relates to the active measurement area for each foot previous work 1 shows that it is acceptable a coverage area of 30 cm x 30 cm on each foot as a result of changing constantly feet positions Then considering the foot typically inside that area and the limited number of 24 sensors available the sensing elements were equally distributed into two distinct areas according the scheme presented on Figure 9 The sensors areas are named rear platform and front platform for identification purposes In this way it is possible to cover an area of 17 12 The sensors chosen distribution tried to ensure that at least one sensor would be activated under the forefoot and another under rearfoot reference rear platform front platform 19 Figure 9 Force Sensors distribution and platform identification Each sensor has 2 pins to connect it to the circuit which must receive 24 wires for each platform Due to the thickness of the sensors 0 203 mm it is important that the cables do not physically interfere on the measurement which means that the wires must be thin Furthermore the cable s current range must acce
41. DEVELOPMENT OF A SYSTEM TO MONITOR SURFER S BALANCE AND PLANTAR PRESSURE AND SURFBOARD S MOVEMENT Daniel Dezan de Bona Instituto Superior de Engenharia do Porto Master in Electronic and Computer Engineering Automation and Systems Department of Electrical Engineering School of Engineering 2014 This report fulfills the requirements contained in form Discipline of Thesis Dissertation of 2 year of Master in Engineering of Electric and Computer Candidate Daniel Dezan de Bona N 1110023 daniel dezan gmail com Supervisor Arcelina Marques PhD mmr i1sep ipp pt Co Supervisor Miguel Velhote Correia PhD mcorreia fe up pt 5 e Instituto Superior de Engenharia do Porto Master in Electronic and Computer Engineering Automation and Systems Department of Electrical Engineering School of Engineering 12 de March de 2014 This work is dedicated to my sister Fernanda Dezan de Bona in memoriam who I really miss Acknowledgments I would like to thank everybody that somehow have contributed to the success of this project which beyond of the professional development reported in this text had impact on my personal life First of all I would like to thank God to be a constant presence in my daily routine My parents Geraldo de Bona and Maria Albertina Dezan de Bona for being the most important professors I ever had and also my brother Eduardo de Bona to show me how important family is Likewise I would like
42. Description Indicates the frame begins Time information composed by Hour Minute Second and Millisecond Composed by 24 separated field with information about force sensors in Newton Pitch angle information in degree Roll angle information in degree Instantaneous acceleration about x axis Instantaneous acceleration about y axis Instantaneous acceleration about z axis Center of Pressure position about x axis Center of Pressure position about y axis Center of Pressure absolute force value 106 LOI APPENDIX I Center of Pressure firmware code 319 328 321 322 323 324 325 325 327 328 329 338 331 332 333 getCoP void 1 int i int j 8 8 i 8 1 N SENSORS i 44 i j uipSensor i calculate the contribution of each sensor for the CoP vector position if 1412 rear contribution 1 dResultantRearX dForce j double sensor i dResultantRearY dForce j double sensor i 1 dResultantRear dForce j calculate individual matrix resultant force else front contribution 1 dFrontX dForce j double sensor i 12 6 452 dResultantFrontY dForce j double sensor i 12 1 dResultantFront dForce j calculate individual matrix resultant force y 901 calculate total resultant force dResultantTotal dResultantRear dResultantFront if dResultantRear gt 8 1 dResultantRearX dResultantRear dResultantRearY dResultantRe
43. LEDGMIUUEN 5 1 RESUMO ABSILRACO Das cT V ENDE eT ete ie UOTE OT VIII INDEX OF EIGURES ntis deiode ta 22 XI INDEX OB TAB ss Das EDI XIII ACRONYMS M XV 1 2222 coule eee c eve 1 1 1 1 ____ _6_ 2 1 2 Spe CAN cM 2 1 3 S CHE DUE Bs 2 1 4 ORGANIZATION 222 3 2 Ea ERU EU OE Fa OX 5 2 1 EN TROD UC TION 5 22 SUREING E HARACIERIZATION S DR UE 7 223 MAJOR SUREING MANOEUNVERS ia ad 8 2 4 COMMON 0101 556 6 0 2 AS 14 2 5 EVOLUHONOPSUREEVMALUATION Umen ME CU ET A 15 2 6 VARIABLES TO MEASURE
44. PI Table 5 Peripherals available in the AVR ATMEGA1280 Peripheral Features Timer Counter Two 8 bit Timer Counter with Prescaler and Compare Mode Four 16 bit Timer Counter with Prescaler Compare and Capture Mode ADC 16 channel 10 bit analog to digital converter PWM Four 8 bit PWM channel Six Twelve PWM channels with Programmable Resolution from 2 to 16 bit USART UART Four Programmable Serial USART UART SPI Master Slave Serial Peripheral Interface DC 2 wire serial interface Watchdog timer Programmable Watchdog Timer with separate on chip oscillator The circuit connections with the microcontroller are shown in APPENDIX C The ADC is connected to the circuit from channels 9 to 14 where the connections labeled as XRATE YRATE XOUT YOUT and ZOUT represents de IMU signals from gyroscope and accelerometer and the connection OUT SENS is the output of the force sensors conditioning circuit The signals from the accelerometer must have a parallel capacitor connected to each output to select the bandwidth limit However the use of a commercial IMU solution eliminates the need of external components because they are already implemented Likewise for the gyroscope the need for a RC circuit 15 provided by the IMU circuit That will be explained afterwards on IMU description To provide an internal voltage reference for the ADC a capacitor of 100 nF 15 connected into pin AREF and an internal circuit sets the reference according to the programmin
45. RTL endTxF BufferUSARTI endTxI verify new buffer data 218 UDRI BufferUSARTI bufferTx BufferUSARTI endTx1 send buffer data 211 while TXC1 waits transmit flag indicate end of transmition 212 if BufferUSARTl endTxl lt BUFTX 2 4 update buffer address 213 BufferUSART1 endTxI Ht 214 else 215 BufferUSARTl endTxlI 8 216 telse 217 UCSRIB 8 1 lt lt UDRIE1 disable interrupt 218 Figure 31 UART transmit code routine 24 4 3 3 SENSORS ACQUISITION The function initADC initialize ADC peripheral feature setting all parameters required for start acquisition properly as described in the firmware design topic Furthermore this function initializes the variables that describe force sensor position in the matrix The conversion result is configured as left adjusted which means that the 8 most significant bit are stored in ADCH meanwhile last 2 bit are stored in ADCL as show Figure 32 15 i4 13 12 11 10 9 8 ADO T ADOS ADC ADE ADOS ADC ADO ADC RU DC CO 7 6 3 2 1 0 4 ADCH ADCL Figure 32 ADC store result sequence Considering that the applied signals in the ADC are always positive and varies from GND to Vsys the voltage reference is configured to Vsys using an external capacitor connected to AREF pin The acquisition routines include control over the multiplexer circuit to select the sensor according to the sweeping
46. Typical Performance Evaluation Conditions nae lt 3 Line drawn from O to 50 load epeatability ip Conditioned sensor 80 of full force applied Conditioned sensor 80 of full force applied Drift o Time AEEA iie amsor to ra an input force Operating Temperature 40 F 140 40 C 60 C Porce reading change per degree a 4m perature change 0 036950 PN Tekscan Inc 307 West First Street South Boston 02127 1309 USA tel 17 484 4500 7800 248 3669 5 fax 817 464 1288 e mail marketing com URL E 8 30 13 87 99 APPENDIX Circuit schematic diagram Communication devices and micro SD connector SENS 09268 IMU connections 68 FID gn 1 hino O E I ANTES a Pi TEEN x AREE ZEL 06 T F EE TTE 91 i 91 CO 1 1 ie __ mA e e a v6 APPENDIX D SENS 09268 IMU Schematic diagram Accelerometer TITLE w 25 Document Number Date 9 6 2889 1 57 44 PM Sheet 1 1 56 APPENDIX E Printed Circuit Board PCB Top Layer and Bottom Layer scaled in 2 1 9
47. XD and FTD TXD are the data connections with the microcontroller UARTO the signals D 33 and D connect the FTDI to the USB connector the signals LED_RX and LED TX for LED connection to provide visual information about traffic data and the power source Vsys and GND Afterwards the microcontroller should send the data for the FTDI232R through the FTD pin and should receive the data over the FTD RXD The USB converter data transfer baud rate could achieve 3 Mbps however the maximum baud rate used was 1 Mbps limited by the microcontroller JMP RESET EN U7 R10 DIR 2 TXD OSCO RESET 100nF 5108 E t DC 2 Dm 1 54 D 9 1 CBUS2 boss 543 3 cBUS3 VSBDP FT232RL C17 Figure 15 FTDD32R schematic connection Like the USB converter the WiFi module communicates with the microcontroller through the UART channel however they are independent each other The WiFi module 15 connected to the UARTI channel by pins WiFi RX and WiFi TX and no flux control was implemented The LED connections allow a visual feedback of the module status STATUS data transfer DATA TRAN and module connection CONNEC The jumper JP5 and the pull up resistor R25 enable the system for work in ad hoc mode immediately after being powered Figure 16 shows the RN 131 connections STATUS DATA_TRAN 35 34 33 gt lt 32 31 30 x 29 28 27 et pt t
48. a qualitative analysis it is more difficult to measure the physical variables involved like force angles and speed for example 11 Table 1 Set of manouvers listed and the controlling factors to execute it Manoeuver Upper body Carving Re entry Cut back Floater Body and arms opened in the beginning moving arms down when reaches top wave Follow trunk rotation Body and arms opened in the beginning moving arms down when reaches top wave Keep body position centered during riding to control the balance moving down when speed get slow Factors Trunk rotation Slight rotation Strong rotation almost 180 Slight rotation the beginning Strong rotation in the end when reaches the wave s lip Slight rotation CoP displacement Starts moving the weight forward to reach the top of the wave Then move the weight backward to make pressure on the back foot Starts moving the weight forward to reach the top of the wave Then move the weight backward to make pressure on the back foot to complete the turn After it move weight forward to take off Starts moving the weight forward to reach the top of the wave Then move the weight backward to make pressure on the back foot to complete the turn After it move weight forward to reach the wave s lip Then move the weight backward to make pressure on the back foot to complete the turn Most part of time the center of pres
49. a potential factor to surfers getting injured Injuries are the main cause of surfer absence from competition the ASP website reports almost eleven injuries from 2011 to 2012 on professional men s ranking Apart from that during the 2013 eleven more injured surfers were registered leaving them out at least one event out of 10 events The worst reported case is the absence from 8 events 22 In these reported cases most surfers developed the injuries during training sessions while only a few happened during competitions The complex movement during surfing raises questions on the surfer s vulnerability to injuries Although the most frequent type of injuries is lacerations 41 46 of all surfing injuries 9 there are several injuries related to overextending Mainly at the professional level musculoskeletal injuries are another type of common injury 6 18 Knee stress bad posture and again overextending generate the right scenario for injury to occur Considering a paddling time of 44 51 12 16 surfers can be more susceptible to injuries due to muscular fatigue for example Nathanson et al 7 reported that 62 of the surfing injuries occur during wave riding Although most of them could be connecting with excessive use of the upper body during paddling time like shoulder dislocation and shoulder strain that represents 35 of the total upper body injuries The author reaches that 37 of the total acute injuries occurs o
50. ad used in calibration 1s shown in Table 2 24 Table 2 Force range used in 5 calibration Load g Force N 59 62 0 58 109 62 1 08 159 62 1 57 209 62 2 06 259 62 2 55 309 62 3 04 359 62 3 53 409 62 4 02 438 62 4 30 First of all the loads were applied to the sensor and the voltage measured on the output of the conditioning circuit by using a digital oscilloscope The results have shown that all sensors presented a coefficient of correlation R2 higher than 0 98 with one exception of Sensor 21 that presented 0 93 between voltage and force The Table 3 shows the R calculated Table 3 Coefficient of correlation of 4 4 N force range sensors Sensor Seno R 3 0995160 15 099477 4 0 990279 16 0 992899 5 0 988782 17 0 986208 6 0 996928 18 0 997648 7 0 989282 19 0 999469 8 0 999411 20 0 997267 9 0 989898 2 0 939296 10 0 996232 22 0 998948 11 0 998698 23 0 991836 12 0 991911 24 0 988847 To illustrate the first test response Figure 11 depicts the data from Sensor 8 which present the highest Due to the linearity of the voltage response the equation coefficients were 25 calculated for each sensor and placed in the ADC conversion function in the firmware as will be explained later KQ 1800 1600 1400 1200 1000 800 600 400 200 Figure 11 Voltage and resistance response for Sensor 8 Resistance vs Force Voltage vs Force R
51. als and some functions like multiplexer control Figure 22 shows the sequence of initialization Setup Multiplexer init ADC init Calibration Interrupts enable UART init SPI Timers init End Figure 22 Setup sequence for system initialization At first the multiplexer 1s initialized by assigning proper I O pins direction and signals state as described in Table 9 This configuration ensures that the multiplexers are disabled After that the global interrupt enable command 15 executed due to every interrupt vector are dependent of the global interrupt state Once the global interrupt 1s enabled every interrupt can be controlled individually by its own register 42 Table 9 Multiplexer pinout configuration Pin Description Configuration State PAO channel selection A output Low PA1 channel selection B output Low 2 channel selection output Low PA3 mux enable 1 output High PA4 mux enable 2 output High 5 3 output High PA6 mux enable 4 output High PA7 mux enable 5 output High PD5 channel selection A out output Low PD6 channel selection B out output Low PD7 channel selection out output Low The next setup step 1s the assignment of the UART peripheral registers This configuration sets the serial baud rate frame format and enable the transmitting receiving data Due to the usage of two UART channels the configurations must be done for each one independently and can assume differe
52. ar else dResultantRearX 8 dResultantRearY 8 H if dResultantFront gt dFrontX dResultantFront dResultantFrontY dResultantFront else 1 dFrontX 6 dResultantFrontY 8 1 if dResultantTotal gt H calculate CoP positioning from individual contribution of both matrices dResultantTotalX dFrontX dResultantFront dResultantRearX dResultantRear dResultantTotal dResultantTotalY dResultantFrontY dResultantFront dResultantRearY dResultantRear dResultantTotal dResultantRear 8 dResultantFront 68 dResultantRearX 68 dResultantRearY 8 dFrontX 6 dResultantFrontY 8 15 dResultantTotalX 45 dResultantTotalY 18 dResultantTotal 6 APPENDIX J Matlab code clear all clean workspace amp variables cle system Initializatioh eee fclose instrfind Sclose serial channel if open s serial COM8 BaudRate 115200 Sopen serial channel WiFi tcpip 169 254 1 1 2000 Sopen WiFi channel fopen WiFi open file WiFi s open serial file file fopen C Users Zimba Desktop Teste BOSU 4 calibrado balancel txt w Sopen local file to record frame data colormap jet color map pattern for charts _ esse 0 snsc SS SSS SS SS SS SS Sea 5 linhas 0 number of lines n linhas 400 number of frames to receive n divis es 100 Saux to average linhas anterior 0 Saux to
53. balance feet positioning and surfboard s movement These factors led to the development of a system that 15 capable to measure them dynamically trough measuring plantar forces and surfboard s rotation Beyond measuring these factors the system 15 able to store the data locally trough a mass storage device and send them through WiFi allowing the visualization of the Center of Pressure surfboard s rotation and sensors activation in real time The device is an electronic embedded system composed by a microcontroller ATMEGA1280 a analog conditioning and acquisition circuit an inertial measurement unit a WiFi module RN131 and two matrix of Flexiforce force sensors The embedded firmware was developed in C Language The software Matlab was used to receive and show the data in graphics in real time The tests have shown that the system works properly fulfilling the project requirements providing information about balance through the center of pressure feet positioning through the plantar force distribution and surfboard s movement on pitch and roll axes through the inertial measurement unit The force mean error of measurement of force verified was 0 002 0 0064 N meanwhile the shorter data transmission distance was 100 m The measured power in operation was 330 mW Keywords Center of Pressure Centre of Pressure CoP surfboard instrumentation remote surfing assessment surfe pitch roll vi vii Index ACKNOW
54. bten o de dados biomec nicos que podem ser utilizados tanto no treinamento quando na melhoria da qualidade de vida dos atletas auxiliando na preven o de les es por exemplo Deste modo o presente projeto se aplica na rea do desporto nomeadamente na modalidade do surfe onde a aus ncia de trabalhos cient ficos ainda elevada aliando a tecnologia eletr nica ao desporto para quantificar informa es at ent o desconhecidas Tr s fatores b sicos de desempenho foram levantados sendo eles equil brio posicionamento dos p s e movimenta o da prancha de surfe Estes fatores levaram ao desenvolvimento de um sistema capaz de medi los dinamicamente atrav s da medi o das for as plantares e da rota o da prancha de surfe Al m da medi o dos fatores o sistema capaz de armazenar os dados adquiridos localmente atrav s de um cart o de mem ria para posterior an lise e tamb m envi los atrav s de uma comunica o sem fio permitindo a visualiza o do centro de press es plantares dos ngulos de rota o da prancha de surfe e da ativa o dos sensores em tempo real O dispositivo consiste em um sistema eletr nico embarcado composto por um microcontrolador ATMEGA 1280 um circuito de aquisi o e condicionamento de sinal anal gico uma central inercial um m dulo de comunica o sem fio RN131 um conjunto de sensores de for a Flexiforce O firmware embarcado foi desenvolvido em linguagem O software Ma
55. caler is used providing 128 counts per second The flowchart is shown in APPENDIX F When an interrupt occurs in Timer 2 vector the global RTC variables are updated controlling the RTC operation Furthermore the function monitors the occurrence of leap years Figure 29 shows a code extract for time update 50 61 TMR2 overflow interrupt vector 62 ISR TIMER2 OVF vect 75 to verify that time date is not been updated 76 if lupDateTimeAndDate 77 78 if ucMili 128 79 80 ucSec 1 update seconds 81 ucMili 8 83 if ucSec 68 84 85 ucMin 1 update minute 66 ucSec 0 88 if ucMin 60 89 90 ucHou 1 update hour 91 ucMin 8 93 if ucHou 24 94 1 95 ucHou 8 1 update day 99 180 1 181 else ucMili 1 update milliseconds Figure 29 Timer 2 counter interrupt routine for RTC control To control data communication an UART vector interrupt is used Due to two UART channels implemented UARTO and UARTI on this firmware description will be used only UART to represent both channels There are two interrupt vectors for UART communication controlling receive and transmit separately The data received transmitted by UART channels are stored into UDR register which is a 16 bit where the upper byte 1s the receive byte RXB and the lower byte 15 the transmit byte TXB The interrupt for UART receive occurs when the regis
56. celeration and gyro unit conversion 63 Pieure 45 Pitch and Roll ode PDD Fa tous 64 Figure 46 Matlab application viewing CoP and position information 65 Figure 47 Unstable setup to perform simulation tests and axis orientation 67 Figure 48 Hardware test sep ausencia vie wien E I eu uuu ubl 68 ADC frequency ver E 69 Figure 50 Multiplexer enable timing verification eene eene 70 Figure 51 Multiplexers counting channel selection eere 70 Figure 52 Sensor signal width and fall iere Giese aie 71 shi range distal CE eS da pu Te quintus hd He 72 Figure 54 CoP displacement over the surfboard during tests 73 Figure 55 Feet position from plantar pressure measurement 74 Figure 56 Pitch and Roll along the Test 2 and Test 3 75 Lisure 27 Piten and Roll along Test a 76 xii Index of Tables Table 1 Set of manouvers listed and the controlling factors to execute 12 Table 2 Force range used in sensor s calibration
57. cited 2013 15 05 Available from http www surffcs com 6 Everline C Shortboard Performance Surfing A Qualitative Assessment of Maneuvers and a Sample Periodized Strength and Conditioning Program In and Out of the Water Strength amp Conditioning Journal 2007 29 3 p 32 40 7 Nathanson Haynes and D Galanis Surfing injuries The American Journal of Emergency Medicine 2002 20 3 p 155 160 8 Nathanson A et al Competitive Surfing Injuries A Prospective Study of Surfing Related Injuries Among Contest Surfers The American Journal of Sports Medicine 2007 35 1 p 113 117 9 Lowdon B N Pateman and A Pitman Surfboard riding injuries The Medical Journal of Australia 1983 2 12 p 613 81 10 Taylor D M et al Acute injury and chronic disability resulting from surfboard riding Journal of science and medicine in sport Sports Medicine Australia 2004 7 4 p 429 437 11 Farley O R L Harris and Kilding Physiological Demands of Competitive Surfing The Journal of Strength amp Conditioning Research 2012 26 7 p 1887 1896 12 R A Lowdon and Davie Heart rates and estimated energy expenditure during recreational surfing Australian Journal of Science and Medicine in Sport 1991 23 3 p 70 74 13 Loveless DJ and C Minahan Two reliable protocols for assessing maximal paddling performance in surfboard riders Journal
58. d disable as one logic and enable as zero logic 69 T g 2 0065 5 _ i p oio dl oL eee Measure y ent Mean 35us 20 0B7ms 88 EN 37 887 5 731 Figure 50 Multiplexer enable timing verification 5 1 2 MULTIPLEXER COUNTING SELECTION The counting selection 15 responsible for selecting the multiplexer channel that will be read by the ADC in the microcontroller altogether with the enable signals Due to the multiplexer cascade architecture the counting routine from signals SEL A SEL B and SEL C is performed three times for each ADC acquisition routine and is synchronized with the enable signals as shows in Figure 51 Due to the use of 24 force sensors the signal SEL C OUT has not been used which explain its absence in the screenshot ES A Acquisition 10 0MSa s Do Dio Figure 51 Multiplexers counting channel selection 70 5 1 3 SENSOR SIGNAL ACQUISITION To verify the signals from sensor acquisition the oscilloscope was connected to the conditioning circuit output to measure sensor s signal width and fall time The sensor s signal width is defined as the time between changes in the level of the least significant bit LSB of the multiplexer select counter SEL_A On the other hand the fall time is defined as the accommodation time that occurs after changes in the LSB To measure thes
59. de the frame received into local variables 5 sensor 1 MD tras 4 1 dado 1 3 5 Sensor tras 6 3 sensor tras 6 5 MD tras 6 3 sensor 2 MD tras 3 1 dado 1 4 sensor tras 6 4 sensor Lras 6o 4 1T MD tras 6 4 sensor 3 MD tras 5 2 dado l 6 sensor Lirias 5 2 Sensor Lras o 2 MD tras 5 2 4 sensor 4 MD tras 2 2 dado 1 5 5 Sensor Cras 5 27 Sensor LEBSS 2 9 MD 2 2 sensor 5 MD tras 4 3 dado 1 13 5 sensor tras 4 3 sensor Lras 4 2 9 MD tras 4 3 sensor 6 MD tras 3 3 dado 1 7 Sensor Cras 4 4 sensor tras 4 4 7 MD 4 4 sensor 7 MD 5 4 1 16 sensor Lras 2 2 sensor tras 3 2 7 MD 2 2 lt sensor 8 MD tras 2 4 dado 1 9 5 sensor tras oyo Sensor MD Lras 2 9 3 9 MD tras 4 5 dado 1 17 5 Sensor tras Sensor Lrasi2 2 v MD sensor 10 MD 3 5 1 9 sensor tras 2 4 sensor tras z 4 MD tras 2 4 sensor 11 MD tras 5 6 dado 1 15 sensor sensor 1 2 MD trasi sensor 12 MD 2 6 Gado 1 10 5 sensor tras l o sensor trasll 5 MD tras 1 3 7 sensor 13 MD Frente 4 1 dado l 225 5 Sensor rence 6 3 Sensor Irente 6 5 MD Trente 6G gt sensor 14 MD frente 3 1 1 24 sensor 0 4 Censo 0 4 MD
60. dization the values are converted to Newton assuming the linearity voltage versus force by using the equation from the calibration tests An extract of the complete code found in APPENDIX G is shown in Figure 38 for sensors 6 to 8 57 197 jvoid getForceSensors void 198 4 199 235 m 246 dForce uipSensor 5 2 5768 dSensorValue uipSensor 5 9 2185 sensor 6 247 if dForce uipSensor 5 8 2185 dForce uipSensor 5 8 248 dForce uipSensor 6 2 9821 dSensorValue uipSensor 6 8 3926 sensor 7 249 if dForce uipSensor 6 lt 8 3926 dForce uipSensor 6 8 258 dForce uipSensor 7 2 6214 dSensorValue uipSensor 7 6 3475 sensor 8 251 if dForce uipSensor 7 lt 8 3475 dForce uipSensor 7 252 4 286 Figure 38 ADC result to force conversion On the IMU conversion the conversion results are stored in the variable uiAccelXYZGiroXY as depicted in Figure 39 217 getAccel yro void 218 219 uint8 t i 220 for i 1 i N 14 multiplex N SENSORS 221 222 ADCSelectChannel i select adc channel 223 ADCSRA 1 lt lt ADSC Start next conversion 224 while ADCSRA amp 8x4HB wait end of conversion 225 uiAccelXYzGiroXY i iADCResult storage sensor s conversion 226 227 data sensor 1 228 Figure 39 IMU conversion code routine 4 3 4 DATA STORAGE All data processed from the force sensors position
61. e parameters one sensor was activated and the output is shown in Figure 52 The signal presented width of 334 4us and the rise and fall times of 3 19 us and 3 69 us respectively 1 1 007 2 200V 1 5165 100 057 Trig d 2 750 Measure Current Mean Std Dev Count Ower 1 43 4 Q4 0 64 25 110 Wid 5 9005 pus 1605 Rise 1 3 2 1 9 2us 2 30us 418 agus 25 90 5 464 3 6315 71 849us 2 7505 515 50us 102 03 5 469 Figure 52 Sensor signal width and times 5 1 4 DATA TRANSFER DISTANCE The data transfer distance was verified in open field by sending the data frame through WiFi for a computer increase the distance until the signal gets lost The test was performed three times where the position of the WiFi on chip antenna in relation to the observer on the computer was changed as depicted in Figure 53 Due to the antenna specification the longest distance must be achieved on plane ZY 35 2 The accommodation time is considered in between 90 to 10 of the signal voltage 71 plane YX plane ZY Y 5 plane ZX P ms X y Figure 53 WiFi range distance test The results have shown that the minimum covered distance was 100 meters on the plane ZX and the maximum distance was 230 meters in the plane ZY 5 1 5 POWER CONSUMPTION Due to the highest power consumption occurs whe
62. e apart from updating it The user also can verify the list of files recorded into de micro SD Card Finally the file content can be read or deleted from the memory On the other hand once Acquisition mode 15 selected the system will start to run continuously acquiring data from the sensors and converting it into the desirable variables 4 2 3 MAIN OPERATION When the Acquisition mode is selected the system goes into a loop where the flag of Timer tick is continuously verified to perform a new acquisition Once the flag is activated the firmware executes four different functions where at first data from force sensors and IMU is acquired then it 15 processed to provide information about pitch and roll and CoP Those functions will be described on the following topics Data stored into the variables is fitted into a data frame which will be sent for the micro SD card and WiFi This frame generation updates the timestamp to ensure that its information is correct and equivalent to the current time Finally the data frame is sent by WiFi with the same format as was stored in the memory 4 3 IMPLEMENTATION The firmware was developed using C programming language in the Atmel Studio 6 1 and compiled using the native AVR toolchain available within the program The binary code 46 was downloaded to the microcontroller using the AVR Dragon programmer from Atmel via JTAG which also allows the debugging process The code was divided
63. e for occurring injuries These facts show the relevance of developing a system to collect data during any surf session to assist surfers and coaches in understanding the physical stimuli in which the surfers body and the surfboard are subject helping in designing of specific training practices for example The development of a system to monitor surfer and surfboard s movements was a challenge greatly due to the limited number of scientific work published the scarce of quantitative data from the wave riding movements which was only described qualitatively from a few authors 77 The system proposed in this project aimed to fill this gap by developing an electronic system capable of acquiring dynamic data of the center of pressure and surfboard s position during surf practice Although the evaluation in the water has not been possible due to time and budget limitations the tests results have shown that is possible to determine the proposed variables Furthermore the use of a video analysis altogether with this system can contribute to achieve better results due to the visual feedback provided by the images The Flexiforce sensors used to acquire plantar pressure proved to be a good choice for this application due to the response linearity when subject to the force within its operational range The sensors are also lightweight and thin which are relevant characteristics due to the low weight of the surfboards However the terminal connec
64. e heart rate data velocity and time spent in each category of movement The result for that particular situation shows that the average covered distance per heat was 1 605 313 5 m with an average speed of 33 4 6 5 km h 11 Otherwise studies like Mendez Villanueva et al e Loveless et al try to evaluate how the surfers use their technique during the surf session analyzing paddle movement 13 16 Mendez Villanueva et al also evaluated the profile of men s competitive surfing during an international competition To figure out the time expend during surfing four distinct movements categories were analyzed as follows wave ridding stationary and miscellaneous Each category had had the percentage of 51 42 5 3 8 and 2 of the total surfing time respectively In contrast et al obtained percentages of 44 and 35 for paddling and stationary categories in recreational surfing and 5 for wave ridding 19 These results show the gap existing in time spent on each category in different purpose practices With 206 waves the average length of riding wave was 11 6 seconds 16 In contrast Lowdon et al reported higher numbers in between 20 0 and 23 7 seconds both in international contests 20 These differences are justified by the competitions taken place in different locations and surf conditions Despite to be the major goal of surf wave ridding occupies only a small period of 3 8 5 of the t
65. e sensor s matrix available in the market that can provide higher resolution Another future work is the implementation of a supply management circuit that uses the USB connection to charges the battery The integrated circuit bq24278 is one option once it controls the battery charge besides provide control signals which the system can use to determine system capacity for instance Afterwards testing the system in a real situation inside the water during wave riding it is a need for future works characterizing the surf movements allowing comparing different rides and evaluating them 79 80 References 1 Dezan de Bona D G de Salvador Ferreira and L Schwarz Sensoriamento remoto em pranchas de surfe utilizando tecnologia ZigBee INDUSCON 2010 9th IEEE IAS International Conference on Industry Application Sao Paulo Brazil 2010 2 Silva A S et al WIMU WEARABLE INERTIAL MONITORING UNIT BIODEVICES International Conference on Biomedical Electronics and Devices Rome Italy 2011 3 LIEBERMANN Dario G L K T HUGHES Mike D BARTLETT Roger M McCLEMENT Jim S FRANKS Ian M Advances in the application of information technology to spor t performance Journal of Sports Sciences 2002 p 14 4 Gesser F J Possobon F R Bonacorso N G Silva R Desenvolvimento e constru o de uma fresadora CNC de baixo custo destinada a confec o de pranchas de surf 2007 5 Ltd F C S P 2013
66. e sensors activation in a graphical interface Furthermore pitch and roll angles are displayed on command window Figure 46 if firstSample w 6 w lt 23 wtt RwEst w RwAcc w initialize with accelerometer readings else evaluate RwGyro vector if abs RwEst 2 lt 6 151 Rz is too small and because it is used as reference for computing Axz Ayz it s error fluctuations will amplify leading to bad results in this case skip the gyro data and just use previous estimate 0 lt 2 09 RwGyro w RwEst w else i get angles between projection of R on ZX ZY plane and 2 axis based on last RwEst fTor w 8 w amp 1 w 4H 1 Awz w atan2 RwEst w RwEst 2 189 M PI get angle and convert to degrees Awz dRxGyro Gyrodt get updated angle according to gyro movement Awz 1 dRyGyro Gyrodt get updated angle according to gyro movement estimate sign of RzGyro by looking in what quadrant the angle is RzGyro is positive if in range 98 98 gt cos Awz gt 8 signRzGyro cos Awz M PI 188 gt 8 2 1 1 reverse calculation of from angles for w 80 w amp 1 w4H 1 RwGyro sin Awz M PI 180 Rwayro 6 sqrt 1 squared cos Awz 8 M PI 188 squared tan Awz 1 M PI 1860 RwGyro 1 sin Awz 1 M PI 180 RwGyro 1 sqrt 1 squared cos Awz 1 M PI 188 squared tan Awz 6 M 180 1 RwGyro 2 signRzGyro
67. e tasks to be performed in the firmware are e Controlling the multiplexer cascade circuit e Acquiring conditioning circuit and IMU voltages Storing the data from conversions and calculations Transmitting data trough wireless and USB to a remote computer e Receiving wireless and USB data from a remote computer Based on these requirements the next steps define the general firmware structure 4 2 DESIGN The firmware design is described here by using flow diagram The Figure 21 shows the general functionality of the firmware and describes the main flow Initialization menu New data acquisition 5 Figure 21 General firmware functionality 4 Primarily the system initialization routine sets up the peripherals and variables to allow proper operation This setup is called only once during power on After the setup the system enters in an initialization menu where the user can adjust a few parameters such as date and time e g and start the data acquisition Then the system enters in an infinity loop where the data acquisition is performed and the data frame is sent over WiFi Next topics will describe the functionality of each stage in the main flow 4 2 1 SETUP The setup routine initializes the system by setting the microcontroller s register properly according to the hardware architecture and system requirements The routine must initialize the peripher
68. ease performance and help understanding the mechanisms of injuries which is a concern of professionals This project aims to develop an electronic system to be used in action sports using a board by acquiring data from force sensors accelerometer and gyroscope to provide information such as feet positioning CoP displacement and board s rotation Surf was the sport chosen to be the application environment due to the familiarity that the candidate has with this sport making it easier to understand the specific needs 1 Board is general in this context which can be applied as surfboard skateboard snowboard ski etc 1 1 BACKGROUND The idea behind this project follows from a previous work 1 where some changes were identified as a needed improvement on the technology used namely in the sensors applied in that situation to collect plantar pressure In addition no quantitative measurement was presented neither relating feet positioning nor plantar pressures involved in wave riding Following those suggestions and the state of the art analysis this project also includes an inertial measurement unit allowing to determine the surfboard s movement altogether with the plantar pressure distribution 1 2 OBJECTIVE This project aims to develop a system which will allow to measure and collect biomechanical parameters of a surf practitioner while surfing Therefore the system has to provide local backup and wireless data transm
69. ents The system developed overcomes the absence of quantitative data in this field by collecting dynamic data from the plantar pressure and surfboard s rotation to quantify 78 factors known to affect 5 performance like feet position CoP and surfboard s movement Furthermore since the system enables to identify biomechanical parameters that can be used to help in injuries preventions could be a possible development of this system Moreover the system has become the first electronic embedded equipment exclusively applied on surf that had presented results that have shown these factors Finally the project s development contributed to develop ideas on that specific application providing a constant growth in partnerships and also in know hall about it 6 1 FUTURE WORKS The most important future work in this project is to integrate the electronic system into a waterproof case to allow the use during real situation One possible solution consists in the use of a waterproof case for the PCB and a rubber or silicon sheet to seal the sensors area However other solutions can be tested like the use of flexible electronic circuits to allow the integration between sensors and electronics removing the need of cables and reducing the equipment size To improve the system is also important mainly in the establishment of the feet positioning it is an advised to increase the number of sensors which can be done by using th
70. form due to the pitch rotation performed by the surfer Moreover in Test 3 is observed the highest rotation deviation mainly in Roll axis which corroborates with the movements performed to move the surfboard sideways Therewith the test results have shown that the system is able to measure and quantify the CoP displacement Figure 54 shows the CoP displacement over the surfboard drawing where 15 clearly noticeable the CoP displacement backward during Test 2 and the highest deviation during Test 3 cm CoP Displacement CoP Test 1 CoP Test 2 X CoP Test3 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 cm Figure 54 CoP displacement over the surfboard during tests performed 73 5 2 2 FEET POSITION To determine the feet position the plantar pressure measurement is done by reading the force sensors and due to the knowledge of the surfer s foot stance it is possible to establish the position Due to the data measured from the force sensors and the analysis of the sensors that presented an average value greater than zero the foot position was establish for each trial The Figure 55 shows the suggested position in one trial of each test Figure 55 Feet position from plantar pressure measurement 74 It was observed during these tests that due to the low spatial resolution it was difficult to establish the feet position accurately meaning that the spatial resolutio
71. g Research 2010 24 1 p 135 139 23 Association of Surfing Professionals ASP World Championship Tour Ranking 2013 cited 2014 03 01 Available from http www aspworldtour com rankings asp world championship tour ranking 24 Foundation T R I Pukas and Tecnalia work together to develop the world s first surfboard with integrated techology 2011 22 02 2011 cited 2011 10 03 Available from http www tecnalia com en press room press releases tecnalia pukas tecnalia surfboard integrated technology htm 25 AlpineReplay Trace ActiveReplay 2013 cited 2013 10 10 Available from http www alpinereplay com trace pre order 26 Winter D A A E Patla and J S Frank Assessment of balance control in humans Med Prog Technol 1990 16 1 2 p 31 51 27 L M Practical biomechanics and physiology of balance Handbook of balance function and testing 1993 p 261 279 28 Hrysomallis C Balance ability and athletic performance Sports Medicine 2011 41 3 p 221 232 29 Incorporation T FlexiForce Sensor User Manual 2010 15 30 Devices A ADXL335 Small Low Power 3 Axis 3 Accelerometer 2010 cited 2013 17 11 Available from http www analog com staticimported files data sheets ADXL335 pdf 31 Incorporation I IDG 500 Dual Axis Gyro Product Specification 2009 cited 2013 03 04 Available from http Anvensense com mems gyro documents PS IDG 0500 00 06
72. g code Another connection to the microcontroller is the WiFi module This module communicates using an UART 28 protocol The pins WiFi RX and WiFi_TX from the module are connected to the microcontroller UARTI on Port D RXD1 and TXD1 To generate a time reference the Timer Counter 2 is used in a RTC configuration In that configuration it is recommended the use of an external oscillator of 32 768 kHz because this timer is optimized for better accuracy when in RTC mode Due to that a crystal X2 18 connected to the microcontroller through the pins TOSCI TOSC2 Two capacitors of 6 pF are connected to the crystal and GND to warrantee the accuracy specified by the manufacturer The main frequency is generated by a 16 MHz crystal connected to the and XTAL2 pins on the microcontroller which is the maximum clock frequency allowed for this device Two 22 pF capacitors are connected to the crystal and GND to avoid unwanted oscillations and noise A general external reset circuit is implemented using a RC configuration The reset circuit is low active and the minimum pulse period to reset the microcontroller is 2 5 us The control of the multiplexed conditioning circuit 15 done by eight pins from Port A and three pins from Port D These pins control the multiplexer s output selection and multiplexer enable in order to sweep the inputs from the platform sensors Moreover there are five decoupling capacitors connected on the power
73. ifesavers They also conclude that the relation between heart rate and oxygen consumption reserve during arm paddling is not straightforward suggesting the use of an individual relational equation Furthermore a qualitative description of some of the major surfing movements and manoeuvers identify time expenditure for different activities and surfer movements analyzing surfers during the practice even in professional or recreational sessions 6 15 16 As a professional sport surfing has its own international association in which rules and standards are set for the contests The Association of Surfing Professionals ASP is the responsible agency to managing the ASP Rule Book 17 among other activities like anti doping and ranking control for example The ASP Rule Book is used as a reference for the contests judges to score every surfer s wave according to the parameters listed e http www powerlightsurfboards com br 2 2 SURFING CHARACTERIZATION From the literature it 1s seen that just a small part of the surfing activity 15 spent on wave riding 11 13 16 18 but 1 does not mean that a good physical and physiological condition is not necessary since others factors like wave conditions and psychological stress may also be present Farley et al concerned about physiological demand of competition surfing using a heart rate monitor a global system positioning GPS and a time motion analysis using video to acquir
74. igure 35 17 Derived from International System Units Newton is a unit to quantify force and represents either as N or kg m s 55 Force Sensor Acquisition Select ADC Channel All Sensors Acquired Sets new value flag Select Multiplexer Channel Acquisitions Average n Start ADC Acquisiton Acquisition has finished Add converted value to the average filter Finish Average Filter Sensor Values gt Calibration Set Sensor Value to Zero Subtracts calibration from sensor value Converts Sensor value to force in Newton Figure 35 ADC Force sensors acquisition flowchart This function starts selecting the ADC channel 14 and the conversion starts inside a for loop with the escape parameter N SENSOR Inside this loop the function selectMuxChannel selects the multiplexer input channel as shown in Figure 36 197 void getForceSensors void 198 1 199 El 211 ADCSelectChannel 6 212 for i 8 i N 5 5085 14 213 214 1 2 215 selectMuxChannel i ENABLE 216 T delay us 108 select channel 14 of adc converter multiplex N SENSORS select ADC channel Delay due to the multiplexer stabilization Figure 36 Force sensor acquire extract function code To avoid noise generated by switching delay of 100 us is added in before the conversion starts 56 After tha
75. ing elbows back Figure 1a putting the hands on board Figure 1b and pushing the surfboard down and the body up Figure Ic On the last movement the surfer must control the surfboard to follow the wave flow Figure 14 After take off the surfer is able to ride the wave According Everline the surfer s position in relation to the surfboard is frequently half squatting with flexed knees after take off From this period onwards the movements depend of the leg strength and the balance which allows the surfer control the surfboard direction and movement as depicted in Figure 2a After take off and positioning the surfer executes the bottom turn changing the natural path of the surfboard in direction of the beach executing a turn on the base as depicted in Figure 2b The bottom turn 15 a critical manoeuver that allows the surfer follow the wave wall direction Surfer s foot position will dictate the mechanics of the manoeuvers during and after the bottom turn The surfer changes the balance putting more weight to the surfboard s tail and flexing his knee 6 positioning b bottom turn and positioning Figure 2 Surfer positioning after take off and bottom turn approach There are two foot stances in surf regular foot and goofy foot The difference between them resides on the foot that goes in front Regular footers put the left foot in front while goofy footers put the right foot in front For the regular ones a left
76. ioning Lower Extremities Center of Pressure Inertial Measurement Unit Electrically Erasable Programmable Read Only Memory Static Random Access Memory Reduced Instruction Set Computing Joint Test Action Group Analog to Digital Converter Universal Asynchronous Receiver Transmitter Serial Peripheral Interface Operational Amplifier Printed Circuit Board Surface Mount Device Universal Serial Bus Input Output Real Timer Counter Clock Least significant bit Most significant bit XV XVI 1 INTRODUCTION Nowadays electronic systems have been applied in many distinct areas beyond industrial applications such as home automation car controlling gadgets medicine and even sports The cost reduction of the electronic components and the popularization of technology for daily use also contribute for the development of innovations either for simple mobile phone applications or more complex systems that can control the electric consumption in a house for example Following that trend in sports the electronic systems have been used in different manners e g for obtaining better accuracy in real time acquisition during running competitions However they have also been applied in sports to assist in movement capture and analysis in order to understand the body dynamic behavior Those applications can provide important information for athletes and coaches figuring out the best technique that can be used to incr
77. itting so that the real time information is collected and sent to the remote station where it can be processed To achieve this goal the project was elaborated with a microcontroller platform as the core of the system which 15 capable of acquiring converting and processing the data The goals of the project were pursued into five main topics Find a sensor capable of measuring plantar force e Use of a inertial unit to acquire the surfboard movements Develop an electronic platform to acquire data from force sensors and a Inertial Measurement Unit Develop the embedded firmware Test and validation of the system Writing the report 1 3 SCHEDULE This project was developed under the Master in Electronic and Computer Engineering program of the School of Engineering of Polytechnic of Porto with a limited period of time of one year In order to manage the time usage a schedule was done according to the objective which is presented in APPENDIX A The schedule includes a group of tasks like hardware development research of the state of the art tests and report writing for example 1 4 REPORT ORGANIZATION Chapter 2 presents the state of the art related to measuring biomechanical parameters during surf practice apart from a description of the major parameters and features of surfing Chapter 3 and 4 present project s development divided into its main parts sensors hardware development and firmware Chapter 6 presents
78. l PER Frotal 6 Y m Fnrront x Yrrront Frrear x Yrrear Total Total The code that implements equations system 5 and 6 15 shown in APPENDIX I A for loop is used to sweep all sensors The array variable sensor provides the information about the position of each sensor 4 3 7 ROTATION ESTIMATIVE The surfboard s position is calculated based on the IMU data The file IMU filter contains the functions that are based on the Starlino Electronics and adapted for this project 34 A function called getEstimatedInclination uses the data from ADC conversion to convert acceleration and gyro measurements into pitch and roll angles The function starts converting data units g for acceleration and s for gyro using sensitivity and offset 62 presented by manufacturers as shown in Figure 44 Due to the IMU position being upside down in relation to gravity force on the PCB an inverter factor 1 is used on z axis All accelerations are normalized to 12 185 Elvoid getEstimatedInclination i 186 int w 187 char signRzGyro 188 E 127 convert to acceleration in g unit 128 dRxAcc double uiAccelXYZGiroXY ACCX AXIS 3 3 1823 1 65 8 33 129 dRyAcc double uiAccelXYZGiroXY ACCY AXIS 3 3 1623 1 65 0 33 138 dRzAcc 1 double uiAccelXYZGiroXY ACCZ 5 3 3 1023 1 8 9 33 131 convert to acceleration in s unit 132 dRxGyro double uiAccelXYZGiroXY GYROX AXIS 3 3 162
79. long with the complex movements creates a tough scenario to seek a solution for its characterization Although surfer performance has been evaluated over the years in professional competitions it remains difficult to get a straightforward continuous feedback for surfer and coach about the surfers performance evolution In this chapter a solution to collect data from sensors under surfer s feet and on the surfboard to provide information about surfing movements 15 presented The system is composed by five blocks as depicted in Figure 6 according to their feature First of all the Input block contains the system inputs from the force sensors and IMU and is characterized by converting physical quantities acceleration angular rotation and force into electrical quantities voltage and electrical resistance On the other hand the Output block contains a micro SD Card as the data storage device and stores all information processed to ensure no data is lost when real time communication fails The Processing block conditions the signals from the Input and processes them to obtain the specified variables There is also a Communication block that provides either wireless connection using WiFi or cable connection USB to enable real time operation and system debug 19 Finally in the Software block the real time charting and local data backup is done using Matlab Conditioning circuit gt CPU
80. model A301 was chosen because of a shorter length of the terminal area which allows better control of the element s positioning during assembly of the matrix Figure 7 illustrates the sensor and the datasheet 1s shown in APPENDIX B Trademark of Interlink Electronics 8 Trademark of Tekscan Incorporation 21 14 55 11 Sensing area 25 4 mm 1 in 6mm 25 in Figure 7 Flexiforce A301 model 28 The Flexiforce sensor is a thin film force sensor designed using a flexible printed circuit technology Sensors assemblage is composed by two polyester layers for support and one layer of a pressure sensitive ink in between Each support layers receives a conductive material to allow the electrical contact and an adhesive to glue the layers together The pins are accessible for 2 male pins crimped through the layers The sensitive ink on the sensor is composed by a resistive material that changes the resistance when subject to a force variation The resistance of the unloaded circuit 15 greater than 5 MQ and decreases when force is applied A typical sensor s response for Force vs Resistance Conductance of a 100 Ib range 15 shown in Figure 8 10015 Sensor 1200 0 020 4 0 018 1900 0 016 D a 2 600 0 010 5 0 008 2 400 0 006 E v 0 004 0 002 0 0 000 Force Ibs Figure 8 Typical response for Flexiforce sensors 28 According to the sensor s specification the sensitive area is 0
81. n is insufficient due to sensor discretization However the results show that it is possible to use this method to establish that feature 5 3 SURFBOARD MOVEMENT The characterization of the surfboard movement over X and Y axes is obtained from the IMU data where pitch and roll rotations and acceleration define sideways forward and backward movements The tests have shown a relation between CoP displacement and surfboard rotation for that specific setup although the surfboard angles can vary without CoP changing due to external forces acting over the surfboard The Table 15 shows the relation between CoP and rotation considering the displacement deviation However to analyze the movements described in the set of manoeuvers chosen it might be useful to identify the transitions for instance when there are changes between surfboard s edges on the water The tests performed aimed to verify the capacity of measure the surfboard s rotation angles Due to that Figure 56 shows both rotation axes during different trials where Test 2 represents the second trial on Test 2 likewise Test 3 represents the first trial on Test 3 Surfboard Rotation on Test 2 and Test 3 20 4 CCN N A 0 4 md 1 1 1 qp 1 L 10 4 20 Vo eed 30 Cine apa Sees
82. n the lower extremities LE where foot knee and ankle injuries were the most common The survey reports significant percentage of the total LE injuries on foot 75 were lacerations On the other hand knee injuries such as sprains meniscal tears and dislocations represent 70 of the knee injuries Although they did not specify other kinds of foot injuries beyond lacerations 7 over use of plantar flexion and dorsal flexion could have been responsible for ankle strain related by ASP As a result of excessive body torque 28 of upper body injuries occur while performing maneuvers which was confirmed by 6 Furthermore it is pointed out 7 that up to 1390 of the upper body injuries occur on the trunk whereas 4396 occur the back 3596 on the 14 chest wall the musculoskeletal strain injuries were prevalent although factures were reported 223 EVOLUTION SURF EVALUATION In 2009 Bona et al tried to understand the wave riding characteristics from the surfer s point of view taking into consideration three principal parameters foot positioning balance and distribution of force on the surfboard s decks An electronic system was developed to collect the data from strain gauge sensors on the board and transmit wireless to a remote central The project was finished before water tests because the strain gauges sensors did not work properly after fiber resin reaches total cure Overall the laboratorial tests showed tha
83. n the data 1s transmitting by WiFi the system was powered on in Acquisition mode while the current was verified on the multimeter The total current verified was 100 mA which gives a power consumption of 330 mW due to the supply voltage of 3 3 V 5 2 PLANTAR PRESSURE The plantar pressure measurements allow to determine the CoP and feet position as will be described next 5 2 1 maintain the balance when over the surfboard the surfers must respond to the changes in surfboard position by changing the CoP dynamically The tests described earlier allow the determination of the CoP displacement and the obtained results are shown in Table 15 The first hundred samples were not considered due to the initial transient conditions 72 Table 15 CoP displacement and rotation along the simulation tests Tests CoP X Test 1 63 72 5 54 Test 2 60 73 9 23 Test 3 67 82 3 79 1 1 Pitch Rotation Roll Rotation 1 90 1 31 0 81 0 97 30 03 1 71 3 54 1 90 3 31 2 65 1 91 12 25 The parameters corresponds to mean and standard deviation and the unit are presented in During Test 1 a minimum deviation in rotation is observed keeping the CoP closest to the mean value in comparison with the other tests On the other hand Tests 2 and 3 presented higher deviations in Pitch and Roll changing the CoP position as well Furthermore the Test 2 presented a CoP displacement in direction of the rear plat
84. nal Typical Maximum HIGH to LOW Input to Output 15 ns 30 ns HIGH to LOW Address Selection 150 ns 300 ns LOW to HIGH Input to Output 15 ns 30 ns LOW to HIGH Address Selection 150 ns 300 ns HIGH to OFF Enable to Input Output 120 ns 240 ns OFF to HIGH Enable to Input Output 140 ns 280 ns LOW to OFF Enable to Input Output 145 ns 290 ns OFF to LOW Enable to Input Output 140 ns 280 ns Due to the time delays it is necessary to ensure that the signal acquisition 15 executed when the multiplexer has completed switched Otherwise the acquisition will get the transitions and the signal will do not correspond to the sensors variation In Chapter 4 the embedded firmware managing properly the delays is described 3 4 INERTIAL MEASUREMENT UNIT The IMU gives information about the surfboard s rotation using one accelerometer and one gyroscope Then the movement of the object can be calculated according the acceleration and rotation parameters It is important to note that the position must be calculated using both parameters to provide better results than individual calculation as is described below The IMU used SENS 09268 from Sparkfun Electronics has five degrees of freedom corresponding to three axis of acceleration X Y and Z and two rotation axis pitch and roll The sensing devices inside the IMU are an ADXL335 accelerometer and an 31 IDG500 gyroscope 29 30 Both devices assembled to guaranty the same orientation
85. nd RTC routines are based on Dharmani libraries which were adapted to this project context 33 The functions into the SD routines file controls the data transfer into microcontroller and SD card through the SPI bus The functions into the FAT32 file controls the file system read write process The functions into the RTC routines file controls the RTC update and conversions 4 3 5 DATA COMMUNICATION The communication between the system and the external applications are done using two UART peripherals Both of them are configured and can work independent of each other The peripheral initialization function receives five parameters which configure the baud rate double speed mode data size parity and number of stop bits as depicted in Figure 40 59 57 58 Serial Init Function 59 51 void initializeUART8 long lBaud char cAsyncDoubleSpeed char DataSizeInBits 62 char cParityEVENorODD char cStopBits 63 Figure 40 initialization parameters The UART library has different functions to transmit data according to the data type required and two functions to receive data depending of the desired channel The most important transmit function is called seriallsrOutChar which sends to FIFO TX buffer a char data type updates the pointers and enables the transmitting interrupt An extract of code for UARTO transmission is shown Figure 41 228 Elvoid serialIsrOutChr char data char usart
86. nt values if it 15 necessary Despite of being independent the frame format used is the same and only the baud rate is change to allow faster communication in USB usage Table 10 shows the register and the values assigned in both channels Table 10 UART 0 and UART registers configuration Register Description Configuration UARTO UART1 UBRRnH Baud rate register High Low defines the UBRRnL baud rate Control and status register to enable disable TxEn 1Mpbs 115200 bps UCSRnB Tx and Rx operation RxEn Control and status register to define f ontrol and status register to define frame UCSRnC 5 Stop 1 format Parity ODD represents the UART channels After UART setup the Timers are initialized as well Both Timer 1 and Timer 2 are used for control acquisition frequency and RTC respectively First of all due to the time constraints presented the Timer 1 15 configured to tick new match every 0 1 s providing 43 the frequency of 10 Hz needed for sensors data acquisition On the other hand Timer 2 is configured to tick in counter overflow providing a period of 7 8 ms due to the 8 bit resolution and the frequency operation of 32 678 Hz provided by the external oscillator The Table 11 shows the register configuration for both timers Register TCCR1B TCCR1B TCCR1A OCR1AH OCR1AL TIMSK1 Register ASSR TCCR2B TCCR2A TCCR2A TIMSK2 1 Table 11 Timers 1 and Time
87. ntry Floater Cut back Three sixty Tube and Aerial 15 10 From both studies we can verify that the execution of most manoeuvers is based three controlling factors 1 upper body movements 2 trunk rotation and 3 Center of Pressure CoP displacement The control of these three factors altogether with the surfer s knowledge of the wave will provide the ideal conditions to execute those manoeuvers Table presents the set of manoeuvers listed and the description of the controlling factors According to Lierbermann the feedback information of the movements 15 a major factor to allow systematic correction in sport performance though the knowledge of the goal 15 also necessary to perceive the need to carry out corrections 3 The most common manner to evaluate a surf athlete during the training riding is using a visual sense where the coaches observes the performance of the surfer from the beach giving information to the surfer on the execution of the manoeuvers Nowadays video motion analysis has been used to record the surf session which will be after evaluated by coaches and surfers Furthermore different angles and more than one camera including waterproof cameras have been used to get the largest possible information to evaluate and understand the surfer movements This method has been successful to date however it 1s fully dependent of the coaches knowledge and his perception of how the surfing is done Thus using
88. nts amp 04 10 01 11 214 review 2013 2013 04 11 07 02 8 Writing Report 2013 2014 70d jan 2013 fev 2013 mar 2013 abr 2013 mai 2013 jun 2013 jul 2013 ago 2013 set 2013 out 2013 nov 2013 dez 2013 jan 2014 EB ID Task Name Start Finish Duration fo 01 5 APPENDIX B Flexiforce Datasheet FlexiForce Standard Force amp Load Sensors Modal A301 Physical Properties 0 203 mm 0 008 in 14mm 55 in 25 mm 1 in 9 53 mm 1 0 375 in diameter area gala 25 4 mm Mylar sm mun m 0 1 in ROHS Compliant 25 in Actual size of sensor Length does not include pins please add approximately 6mm 125 in for pin length for a total length of Standard Force Ranges as tested with circuit shown below Force Range Low 0 44 N Circuit Medium 0 25 Ib 111 N High 0 1001 445 AP Cro RAR pese Force Range Adjustments ar In order to measure higher forces apply a lower drive nna voltage 0 5 V 0 10 V etc and reduce the resistance of the feedback SM resistor 1k min To measure lower forces apply a higher drive o da voltage and increase the resistance of the feedback resistor v dw 7 Supply Voltages siau d be ariman teterenco Ros stor ce m 15 Lest 70 100 Sensor Resistance gt load Max cornommecn ded current 2
89. nts describe the action 9 Figure 2 Surfer positioning after take off and bottom turn approach 10 Figure 3 System positioning over the surfboard and surfboard s orientation 17 Figure 4 Matrix of force sensor elements to measure pressure distribution by 17 Figure 5 Surfboard s pitch roll and yaw rotation axis 18 Figure 6 System block 20 blexitorceA90T 29 ara Med Soo Eti pa Ra a 22 Figure 8 Typical response for Flexiforce sensors 1 1 1 22 Figure 9 Force Sensors distribution and platform 23 Figure 10 Front platform with cables assembled 1 24 Figure 11 Voltage and resistance response for Sensor 8 26 Figure 12 Measurement error associated with force 27 Figure 13 Conditioning circuit implementation 30 Figure 14 Inertial Measurement Unit SENS 092
90. nts is the process and materials which are used to manufacture a surfboard where pre shape for the surfboard can be done by CNC machines 4 or the shaper could use an industrial method named thermoforming either with carbon fiber or Kevlar to produce it Like surfboards manufacturing another industry applying good effort to improve the quality of their products 18 fins industry They are using carbon for instance instead plastic providing an extremely hard and lightweight product which increases the drive pivot and hold 5 Also the manufacture of wetsuits with new materials and fabrication techniques appear every year aiming to provide more flexibility comfort and warmth to the surfer Despite all these improvements good equipment is not enough to increase the surfer performance It is necessary hard training inside or outside the water and good physical conditions 6 Most scientific studies about surfing are focused on surfer s injuries and physiological behavior 7 12 which seek to figure out how the surfer s body is affected by the sport technique and which are the body responses to that kind of stimulation which means trying to understand the reaction of the body for some physical external stimuli Apart from these aspects there are some technical studies related to paddle movement 13 14 where they conclude that paddling requires a high energy expenditure by upper body which is similar to that of competitive swimmers and surf l
91. otal surfing time corresponds to this category of movement Take into consideration a surfing session of one hour it means that the surfer 1s able to ride 10 6 3 1 waves per hour Furthermore the environment where surf is practiced involves different kind of conditions for example different wave sizes type of breaker and line up situation 18 These facts could explain the absence of quantitative studies in this sport generating an underexplored research area According to the ASP Rule Book 17 there are some rules and criteria to be followed for the contests to receive the ASP seal This book 15 also used to define the judgment criteria which must be adopted by every judge creating a standard judgment There are five major elements that contribute to increase wave score e Commitment and degree of difficulty e Innovative and progressive manoeuvers e Combination of major manoeuvers e Variety of maneuvers e Speed power and flow The judges use these five elements to score each wave ridden by the surfer The scoring range varies between zero and ten Although the ASP Rule Book do not specify criteria over ride distance and numbers of maneuvers Peirao et al conclude that these two factors have indirect effect on the judgment result in the events analyzed The study reaches a significant correlation between the scores assigned by judges and the judgment criteria indicating that wave riding time and frequency of manoe
92. pdf 83 32 Limited F T D I FT232R USB UART IC 2010 cited 2013 03 04 Available from http www ftdichip com Support Documents DataSheets ICs DS FT232R pdf 33 Incorporation R N RN 131G amp RN 131C 802 11 b g Wireless LAN Module 2012 cited 2013 05 04 Available from http ww 1 microchip com downloads en DeviceDoc rn 13 1 ds v3 2r pdf 34 DharmaniTech microSD ATmega32 Data Logger 2011 cited 2013 01 11 Available from http www dharmanitech com 35 Project S E A guide to using IMU Accelerometer and Gyroscope Devices in embedded Applications 2009 cited 2013 01 11 Available from http www starlino com imu guide html 36 Ltd A Rufa 2 4 GHz SMD Antenna 2008 cited 2013 04 04 2013 Available from http ww 1 microchip com downloads en DeviceDoc Acc Antanova Chip Ant DS pdf 84 82 95 APPENDIX A Project schedule 07 01 27 02 07 01 24 01 um of the art 2013 2013 144 41 3 characterization 18 01 27 02 29d amp variables 2013 2013 Commercial 18 02 27 02 sensors 2013 2013 18 02 22 01 18 02 15 03 Printed Circuit 30 09 08 11 30d Board 2013 2013 17 12 22 01 31 05 31 05 Assembly amp tests 2013 2013 31 05 04 10 2013 2013 Structure 31 05 05 06 definition 2013 2013 06 08 03 10 04 10 04 10 Debug amp tests 2013 2013 04 10 04 11 04 10 01 11 5 Simulation tests 2013 2013 21d 04 11 04 11 6 Result analysis 2013 2013 Adjustme
93. pt routine returns when the ADCH register is read Due to the 10 bit ADC resolution the highest 6 49 bit of iADCResult variable has no affect on the result The conversion code sequence 15 shown in Figure 27 48 ADC interrupt vector 41 EISR ADC vect 42 14 43 uint8 t uiADCLow ADCL 44 iADCResult ADCH 2 uiADCLow 6 18bits 45 Figure 27 ADC read conversion registers sequence To control sensor acquisition frequency a Timer 1 interrupt vector gives a precise period The Timer 1 initialization selects prescaler of 64 and enables a Clear Time on Compare Match operation mode The comparison value gives 10 Hz frequency operation for Timer interrupt In this mode the interrupt occurs when the comparison between timer count and OCRIAH OCRIAL registers matches Then the timer counter is cleared and then counting starts again Inside the interrupt function there is a Boolean Flag tmr flag that signals a new match Figure 28 shows Timer 1 interrupt routine 54 TMR1 match A interrupt vector 55 HISR TIMERI COMPA 56 14 57 tmri flag tmri flag timer 1 flag 58 Figure 28 Timer 1 interrupt routine Another timer interrupt 15 implemented on Timer 2 This interrupt allows asynchronous mode operation which 15 associated with the RTC microcontroller feature The Timer 2 15 an 8 bit timer counter that 15 configured to generate an interrupt when a counter overflow occurs and no pres
94. pt the conditioning circuit specification and must be cheap Owing to these requirements a 24 way flat cable of 0 89 mm thickness and 30 5 mm width was chosen The cable assemblage is shown in Figure 10 23 Figure 10 Rear and Front platforms with cables assembled To connect the sensors to the conditioning circuit a wire to board connector was used providing the electrical contact through soldering terminals 3 1 1 SENSOR S CALIBRATION The sensor s calibration response for load applications was performed by using the protocol suggested by Tekscan 28 In order to obtain the sensor s response and to quantify the measured error two tests were performed for each sensor The tests aim to measure the output voltage from the conditioning circuit and the data converted by using the calibration equations On the first test the voltage output from the conditioning circuit was recorded to verify the sensor s linearity response After that the calibration equations were calculated and inserted on the firmware application On the second test the conversion results were sent through USB for a terminal and the result was recorded to verify the measurement error after calibration Only the sensors of 4 4 N range were calibrated a set of nine loads was used comprising eight loads of 50 g and one load of 28 6 g plus a bar of 9 62 g used to ensure that the force were being applied on the sensitive area of the sensor The total lo
95. quires data from the force sensors to be used as voltage offset calibration Due to the matrix configuration and the multiplexer sweep the no load sensor s voltage can vary between each other which makes necessary to set the offset that will be used as a reference Finally the setup 15 finished by initializing the SPI by setting the mode of operation frame direction e g and enables the peripheral operation as shown in Table 13 Table 13 SPI registers configuration Register Description Configuration State Enables SPI peripheral SPE High SPCR Selects master operation for microcontroller MSTR High Selects the initial clock rate SPR1 0 64 SPSR Defines double speed SPI2X LOW 4 2 2 INITIALIZATION MENU After setup the parameters the system goes into in a menu mode where the user can select Debug mode to configure date and time on the RTC and manage the files in the micro SD Card The user can also select the Acquisition mode where the system will start to run effectively acquiring data To illustrate the mode menu Figure 23 shows a state diagram 45 Initialization finished Debug selection Acquisition selection start continuosly acquisition Debug mode Acquisition mode Exit selection configure date amp time and manege files Figure 23 Modes of operation from main menu Once Debug mode is selected a new menu shows that mode options where the user can see the actual data and tim
96. r 2 registers configuration Timer 1 Description Defines prescaler Sets mode of operation Sets normal mode of operation for pins output OCOA and OCOB Sets the compare value for timer counting compare Enables timer output compares interrupt enable Timer 2 Description Sets Asynchronous mode for external clock signal Sets Asynchronous mode for external 32 KHz crystal Defines prescaler Sets mode of operation Sets normal mode of operation for pins output OCOA and OCOB Enables interrupt overflow n represents the timer number 2 Clear timer on compare match 9 Normal mode counts always up until overflow occurs Configuration CS10 2 WGM13 0 COM1A1 0 COM1B1 0 OCIE1A Configuration EXCLK AS2 CS20 2 WGM23 0 2 1 0 COM2B1 0 TOIE2 State 64 25000 High State Low High No prescaling Normal 0 High Following this stage the ADC is initialized setting parameters like prescaler voltage reference and resolution for instance Due to the time constraints the ADC 15 configured as shown in Table 12 44 Table 12 ADC registers configuration Register Description Configuration State Turn on the ADC peripheral ADEN High ADCSRA Defines ADPS2 0 ate Enables interrupt operation after conversion ADIE High Start first conversion ADSC High DIDRO eee Digital interrupt registers Disable DIDR2 ADMUX Voltage reference selection REFSO 1 AVCC The next step ac
97. record previous number of lines n medias 0 0 to position plot e priar PrI Cs ae Se See Se Sees 5 estore the average value for both platforms sensor media frente divisoes Sensor media tras cell lyn divisoes estore the sum of the values for front platform Sensor frente Zeros th 1 estore the sum of the values for rear platform Sensor tras sensor media total frente zeros T7 T7 Sensor media tras zeros th matrix to store sensor value according to its position MD frente zerosi T r MD tras zeros 7 7 positioning variables pitch 0 roll 0 flag to determinate which charts to be plot plot COP 12 plot decks 1 plot posicao 07 plot media parcial 0 variables to store de CoP result xr total 0 yr total 0 resultante total 0 Saccelxanterior 0 vel 0 o PE E e E O O keyboard wait for the command to run return while 1 main loop 109 frame s Sread frame from file tamanho pacote size frame define irane Size Sonly execute the next code if data frame is complete if tamanho pacote 2 gt 200 af lIinhasen linhas Scheck numbe of received lines if frame 0 inicioSTR frame 1 1 store the frame start char if inicioSTR dado idstrip frame strip number data from the file linhas linhas 1 divi
98. ribe the major manoeuvers and basically all of them are based on the same body movements 6 15 By controlling the balance and feet positioning the surfer is able to change the surfboard s direction and velocity By rotating the upper body the surfer might control the surfboard s direction as well By choosing feet positions the body movements can be controlled Combining all movements the surfer is able to complete the set of major manoeuvers Thus there are three variables that must be measured Balance Feet position e Surfboard s position According to Winter dynamic balance 1s the ability to perform a task maintaining a stable position 25 and it 15 also describe as the process of maintaining the body s center of gravity vertically over the base of support 26 Furthermore from Hrysomallis literature review the CoP displacement is an indication of balance where minimum CoP motion indicates good balance which is normally measured by using force platforms 27 Therefore to measure the CoP on a surfboard one should expect to have some sort of force platform Although the unstable support provided by surfboards and the type of platforms constructions available nowadays a unidirectional measurement of force is proposed on this project using a distributed plantar pressure measurement system Using the same principle to measure plantar pressures feet position determination is a consequence of plantar pressure measurement
99. s given by the SD card and FAT32 routines These routines take approximately 47 3 ms to perform a new data writing in the card Although the number of bytes contributes to increase the time of data writing the time limitation is given by the library implementation Due to the time constraints presented the highest frequency that can be used avoiding data loss is 11 86 Hz However it 1s appropriate to use a frequency lower than that so the acquisition frequency chosen was 10 Hz 3 8 PRINTED CIRCUIT BOARD LAYOUT A printed circuit board was developed from the schematic diagram presented in APPENDIX C using the software Altium Design 13 1 The PCB prototype contains two layers to allowing laboratorial assembly and cost reduction Surface mounting device SMD technology were used to optimize space usage however the footprints selected should allow soldering without industrial machines which was a restriction to consider Most of components were positioned on the top layer although there are two components on the bottom layer The total size of the PCB is 111x62mm and the layers are detailed 18 in APPENDIX E The Figure 19a shows the PCB produced using the first design version On the other hand The Figure 19b shows the 3D model of the second version of the PCB which contains an improvement in multiplexers positioning and in the micro SD Card footprint 38 TOR BM 4s 4 I233C R M Serial Mecos ID 30 02 5477
100. s results and center of pressure are stored into a micro SD Card To access the card a SPI peripheral and a FAT32 file system library are used to allocate the data into a Windows known format First of all the SPI is initialized during the initialization system into setup function A function spi_init 18 called and the parameters are configured as shown in Table 14 Due to the SD card being a slave device the microcontroller is configured as a master device to control SPI access setting high the MSTR bit To configure SPI clock frequency bit SPRO 58 and SPR1 select the main clock division factor and bit SPI2X enables double mode in master device For initialization the division factor is 64 and no double mode 15 used which provides a SPI clock frequency of 250 kHz Table 14 SPI initialization pinout assigned Pin Description Configuration PBO Chip select SS output PB1 Clock frequency CLK output PB2 Master Output Slave Input MOSI output PB3 Master Input Slave Output MISO input To read and write values through SPI peripheral functions SPT receive and SPI transmit are implemented The read write register SPDR stores the content to be transmitted or received and it is single buffered in the transmit direction and double buffered in the receive direction After an SPI peripheral initialization the SD card is also initialized checking the file system and the storage capacity The files SD routines FAT32 a
101. sary unstable support Figure 47 Unstable setup to perform simulation tests and axis orientation For Test 1 the subject stays in balance keeping the surfboard as parallel as possible to the eround plane For Test 2 the subject applies more weight on the back foot moving the surfboard s nose up until reaching 30 degrees on pitch axis keeping that position until the end of the trial Finally on Test 3 the subject keeps the surfboard parallel to the ground plane moving the surfboard from one side to the other along the roll axis by controlling the plantar dorsiflexion movement to reach 15 degrees Five 40s trials were performed at 10 Hz or400 samples After acquisition data was compiled and analyzed on Microsoft Excel To allow a proper auto calibration routine the surfboard was unloaded and the subject waits for the command to start On that moment the surfboard stay over the BOSU inclined about 17 degrees with surfboard s nose touching the ground The calibration routine ends when the system starts to transmit data After this step a signal 1s given to the surfer and data acquisition starts 67 5 1 HARDWARE INTEGRATION TESTS To perform hardware tests the setup depicted in Figure 48 was implemented To ensure constant power supply during the tests the system was powered with an external power supply model 1672 from BK Precision and the voltage and current verified by the Handheld Digital Multimeter U1251B from Agilent
102. sing takes place Microcontrollers are largely used on embedded system due to the number of on chip peripherals available and the range of commercial options Furthermore the use of microcontrollers allows fast prototyping because of the reduced number of external components to work e g memory and peripherals in comparison with processors For this solution an ATMEGA1280 from Atmel Corporation was used The previous knowledge of AVR 8 bit architecture and the compatibility with system requirements were the choice parameters The selected device is a high performance low power 8 bit microcontroller from Atmel Incorporation which is implemented with a RISC architecture that could provide up to 16 MIPS throughput at 16 MHz of operational frequency The ATMEGA1280 has 128 KB of Flash Memory 4 KB of EEPROM which are both programmable through JTAG AVR is trademark of Atmel Incorporation 27 furthermore has 8KB of SRAM Memory Special features can be used to manage the power like six different sleep modes or to detect Brownout for instance There are different peripherals available like timers analog to digital converters serial communication and Real Time Counters RTC for example The list of the main peripherals is shown in Table 5 In the context of this project the used peripherals are an Analog to Digital Converter ADC 3 Timers 2 Universal Asynchronous Receiver Transmitter UART and a Serial Peripheral Interface S
103. sure is forward Obs EI Manoeuver Upper body Three sixty Move Tube Aerial Reverse arms forward in direction of broken wave s lip in the beginning turning completely with surfer in the end Controls the balance However one hand could grab the surfboard s rail to help in surfboard s control or touch the water to increase grip and decrease the speed Move arms forward in direction of broken wave s lip in the beginning One hand could grab the surfboard s rail while in the air Upper body spins around surfboard s normal axis Factors Trunk rotation Strong rotation Executes a completely 360 turn Once in balance no trunk rotation is required Strong rotation Executes a completely 360 turn CoP displacement Starts moving the weight forward to reach the top of the wave Then move the weight backward to make pressure on the back foot to complete the turn After it move weight forward to takes off Starts moving the weight forward to reach the most critical section Then control backward and forward to follow inside wave Starts moving the weight forward to reach the top of the wave When landing controls weight distribution Obs 2 4 COMMON INJURIES Like other radical sports surfing uses nature as the environment to practice which provides uncontrolled situations for the surfer Besides this the complex movements involved on surfing activity can become
104. t another for loop is executed to provide average filter The escape parameter is AVERAGE which is pre defined with value 4 Inside the loop the ADSC bit 15 set high to start the next conversion and a while loop waits to the ADSC bit to be clear by hardware after conversion ends At this moment the new ADC data is available in the variable iADCResult and its value is added to dSensorValue array After the average for loop ends the sum of values stored in dSensorValue is divided by AVERAGE Then this result is compared with the respective pre stored offset Figure 37 shows the acquisition and average extract code 197 Elvoid getForceSensors void 198 199 218 j28 AVERAGE j read ADC channel AVERAGE times to provide a average filter 219 228 ADCSRA 1 lt lt ADSC Start next conversion 221 while ADCSRA amp 8x48 wait end of conversion 222 dSensorValue i iADCResult storage the sensor value adding the last value to do the average 223 1 224 dSensorValue i AVERAGE 225 226 update sensors value to zero if are residuals floating read 227 if dSensorValue i gt 1 228 dSensorValue i dOffSetcalibrate i 229 else 238 dSensorValue i 231 232 dSensorValue i dSensorValue i 3 3 1023 convert adc value to voltage 233 234 E Figure 37 ADC average calculation and off set adjustment After conversion and offset standar
105. t it was possible to obtain the necessary data based on plantar pressure 1 In 2011 two companies from Spain launched a surfboard with an inertial measurement unit strain gauges and GPS which they called as SurfSens Project the first technological surfboard on the world 23 The results were not disclosed and therefore the concept could not be verified and no others news were released since In 2013 a North American company started a campaign to raise funds to support the development of a new sport tracker known as Active Replay The company already developed a mobile application to trace snowboards and skiers using smart phones resources However the new system promises an innovation allowing user to stick the tracker on the surfboards giving information about distance speed size of wave height of air manoeuvers number of turns length of the ride among others features The device stores data that 1s uploaded to a server after use The launch in the market 1s scheduled for March 2014 24 These projects aim to describe the behavior of the surfer and the surfboard during the wave ride though none of them presented until the moment scientific results Once more this could reflect the difficult of quantitative evaluation of the surfing However they prove that there are some efforts in this area either in academia or in the private sector 15 2 6 VARIABLES MEASURE The studies conducted by Everline and Peirao desc
106. ter UDR receives a new data in RXB Due to a FIFO method used in communication process the byte received 1s stored into the final position of local FIFO buffer bufferRx Then the buffer pointers are updated and the interrupt ends Figure 30 shows the UART interrupt code 51 228 8 Receive USART 1 interrupt 221 EIISR USART1 RX vect 222 223 BufferUSART1l bufferRx BufferUSARTl endRxF UDR1 insert data in the buffer 224 225 if BufferUSARTl endRxF lt BUFTX LEN 2 update buffer address 226 227 BufferUSARTL endRxF 228 229 else 238 231 BufferUSARTl endRxF 6 232 H 233 when an overflow occurs overwrite the oldest byte 234 if BufferUSARTl endRxF BufferUSART1 endRxI 235 236 if BufferUSART1 endRxI lt BUFRX LEN 2 237 238 BufferUSART1 endRxIH 239 248 else 241 242 BufferUSART1 endRxl 8 243 244 245 Figure 30 UART receive interrupt code routine Likewise the interrupt vector that controls UART transmit process uses a FIFO buffer bufferTx to provide the data to be transmitted However this interrupt only occurs when a new data is inserted into the FIFO buffer and the UDRIE bit 15 set in the UCSRB register enabling the interrupt The interrupt is disabled when the FIFO buffer is empty Figure 31 shows interrupt transmit code 286 Trasmist USART 1 interrupt 207 EIISR USART1 vect 208 14 289 if BufferUSA
107. tions have presented bad contact after some use which can be resolved by applying matrices off the shelf The IMU used also proved to be suitable for this application since tests have shown a correct response from the device Despite of not be the most up to date device in the market the IMU complied with the aim of measuring the pitch and roll axes however the five degree of freedom have limited the application since yaw axis was not measured For wireless communication a WiFi module was used instead of technologies like Bluetooth for example since the distance covered is greater The module applied proved to be flexible enough to allow the adhoc communication as was implemented and also to allow configuration to be used with routers if necessary Despite of the maximum power consumption described by the manufacture of 700 mW when transmitting which is far from ideal for remote application using batteries as it was the case the system power consumption verified was 330 mW When testing the developed system it was shown that the measurement error is small enough 0 012 0 064 N to be disregarded in this application Also the linearity present in the sensor s calibration showed that the system works properly The movements simulation protocol allowed testing the system dynamically but despite being a controlled environment the unbalanced support provided by the BOSU Ball created an ideal scenario to simulate those movem
108. tlab foi utilizado para de dados visualiza o em tempo real Os testes realizados demostraram que o funcionamento do sistema atende aos requisitos propostos fornecendo informa o acerca do equil brio atrav s do centro de press es do posicionamento dos p s atrav s da distribui o das press es plantares e do movimento da prancha nos eixos pitch e roll atrav s da central inercial O erro m dio de medi o de iii for a verificado foi de 0 0012 0 0064 N enquanto minima dist ncia alcan ada na transmissao sem fios foi de 100 m pot ncia medida do sistema foi de 330 mW Palavras Chave Centro de Press es Plantares CoP instrumenta o de prancha de surfe avalia o no surfe surfe pitch roll Abstract The use of the technology has been increase on last decades in many different fields On sports is not different Everyday new developments came to help athletes improving their performance allowing achieve results never thought Beyond that the use of technology allows the biomechanical data acquisition which can be used for training development or in injuries prevention for instance Thus this project is developed in sport field particularly in surf modality where the absence of scientific studies still present combining electronic technology and sport to quantify information that remaining unknown Three basic factors known as influence surfer s performance were verified
109. uvers affect judgment indirectly 15 2 3 MAJOR SURFING MANOEUVERS During a surf session conditions can change quickly since every wave 15 different For a surfer being able to reach a great wave riding in these dynamic conditions constitutes a challenge Furthermore being the set of manoeuvers quite large surfers are always looking for evolution and performance improvement During training sessions surfers frequently stay 4 5 hours in the water 18 thus requiring high technical and skill abilities to support it 6 21 In 2007 a study was conducted to describe the physiological demands of surfing using qualitative analysis of major manoeuvers According to Everline the primary movements which the surfer must be capable of are paddling and take off where the surfer moves around to get the surf zone and also take the wave It is necessary that the surfers have the skills to know the correct angle of take off and velocity which can change according to wave type Stand up quickly on the correct instant and have the balance are the next step after the take off 6 Figure 1 shows stand up and take off movements moving elbows back b hands position raising the body d finishing take off Figure 1 Take off movement Four different moments describe the action First the surfer needs paddle to reach the speed to follow the wave until the breaking zone After that he starts the take off movement mov
110. yzing the signals obtained and the system s operation when simulating basic movements present in surfing practice The tests also aim to perform the last two steps presented in the V Model for the firmware development which are hardware integration and validation To verify the hardware integration the electronic circuit was connected to a DSO X 3012A Infini Vision oscilloscope from Agilent Technologies to acquire the signals to be compared with predefined parameters such as delays and acquisition frequency for example Due to the importance for the data acquisition and the number of force sensors to read 24 the multiplexer signals are shown in detail other features such as data transmission and power consumption during transmission for instance were also tested To simulate the movements from surfing activity a test protocol was established using a regular footer surfer to perform three different setups over the surfboard in which Test 1 Balance Test 2 Pitch rotation and Test 3 Roll rotation plantar dorsiflexion sideways displacement Following the description in Chapter 2 of the surfing conditions and context the surfboard can be considered an unstable platform where the surfer stands up over it controlling the balance to perform the manoeuvers as intended Therefore these conditions 66 were simulated by using a BOSU Ball from BOSU Fitness LLC as depicted in Figure 47 which in turn provides the neces

of Figures

Contents

Download Pdf Manuals

Related Search

Related Contents