Home

1Y1`

1. 191 to 19 acquired by the devices 121 to 12n 0033 Furthermore the input control unit 2 inputs BAS data 20 of the BAS 13 and EMS data 21 of the EMS 14 and controls the storage device 4 to store these data 0034 Theimage analysis unit 7 executes analysis process ing for the image data 16 infrared image data 17 and laser image data 18 stored in the storage device 4 generates analy sis data 22 including human information 22a and environ ment information 22b and stores the analysis data 22 in the storage device 4 The image analysis unit 7 implements func tions as a human information generation unit 7a and environ ment information generation unit 76 0035 The human information generation unit 7a extracts feature amounts from the image data 16 infrared image data 17 and laser image data 18 and executes recognition pro cessing and the like based on the extracted feature amounts and set criteria thereby generating the human information 22a 0036 The human information 22a includes the presence absence of a person the number of persons a distribution density of the persons an amount of activity of the person an amount of clothing of the person a personal attribute name gender body type body height age etc a position of the person standing position seated position etc and an activ ity state during office work transfer conversation and so forth of a person in the energy demand prediction area Sep
2. an example of an environment model generation unit and environment analysis unit according to a second embodiment 0015 FIG 4isa block diagram showing an example of a human model generation unit and human analysis unit according to a third embodiment and 0016 FIG 5 is a block diagram showing an example of an arrangement of an energy demand prediction apparatus according to a fourth embodiment DETAILED DESCRIPTION OF THE INVENTION 0017 Embodiments of the present invention will be described hereinafter with reference to the drawings Note that the same reference numerals denote the same or nearly the same components throughout the accompanying draw ings a description thereof will not be given or a brief descrip tion thereof will be given and only differences will be described in detail First Embodiment 0018 This embodiment will explain an energy demand prediction apparatus which analyzes image data acquired by an image sensor to generate image analysis data including at least one of human information and environment informa tion which dynamically change in an energy demand predic tion area and executes an energy demand prediction based on the image analysis data Furthermore in this embodiment the energy demand prediction may be executed further using for example information such as a temperature humidity weather schedule electric power usages and the like in addi tion to the image analysis data 0019 The
3. data by execut ing an energy demand prediction based on the analysis data and an energy demand prediction model generated using previous data corresponding to the analysis data 2 The energy demand prediction apparatus of claim 1 wherein the previous data includes previous data acquired by a device sensor and the prediction unit executes the energy demand prediction based on the analysis data data acquired by the device sensor and the energy demand prediction model 3 The energy demand prediction apparatus of claim 1 further comprising a device control unit that generates based on the human information including a identification result of a person and control setting data including a control value set for the person or an attribute of the person control data which matches the person indicated by the human infor mation and an output control unit that outputs the control data to a corresponding device 4 The energy demand prediction apparatus of claim 1 wherein the environment information includes at least one of light information device layout information and weather information and the energy demand prediction apparatus further comprises an environment model generation unit that generates an environment model which represents a feature of an environment of the prediction target area based on the environment information 5 The energy demand prediction apparatus of claim 1 wherein the human information includes at le
4. energy demand prediction apparatus according to this embodiment accurately measures for example data such as environment information electric energy and human activity information in an energy demand prediction area by analyzing the image data thus attaining an accurate energy demand prediction 0020 FIG 1 is a block diagram showing an example of an arrangement of the energy demand prediction apparatus according to this embodiment 0021 An energy demand prediction apparatus 1 includes an input control unit 2 processor 3 storage device 4 and output control unit 5 0022 The processor 3 functions as an image analysis unit 7 prediction unit 8 and device control unit 24 by executing a program 6 stored in the storage device 4 Note that the image analysis unit 7 prediction unit 8 and device control unit 24 may be implemented by hardware in the energy demand prediction apparatus 1 US 2012 0239213 Al 0023 The energy demand prediction apparatus 1 is con nected to an image sensor 9 infrared sensor 10 laser sensor 11 measurement devices 121 to 127 building automation system BAS building monitoring system 13 and environ ment management system EMS 14 to be able to receive various data associated with an energy demand prediction area Furthermore the energy demand prediction apparatus 1 is connected to an output device 15 and control target devices 251 to 25m 0024 The image sensor 9 includes for example a camera
5. energy saving creation and storage can be attained 0044 Also the device control unit 24 may generate the control data 27 for the devices 251 to 25m based on the human information 22a in the analysis data 22 and the control set ting data 26 including the user information attribute data individual comfort state information and control values cor responding to personal actions which are stored in the stor age device 4 0045 The device control unit 24 identifies an individual based on the human information 22a and generates the con trol data 27 for practicing a comfort state set for this indi US 2012 0239213 Al vidual in the control setting data 26 For example the device control unit 24 may generate the control data 27 for practicing control set for the action state in the control setting data 26 based on the human action state an action sensitive to heat that sensitive to the cold during a desk work during stand talking during walking included in the human information 22a 0046 The output control unit 5 outputs the prediction data 23 and various other data stored in the storage device 4 to the output device 15 0047 Furthermore the output control unit 5 outputs the control data 27 stored in the storage device 4 to the devices 251 to 25m 0048 The output device 15 includes for example a dis play device audio output device communication device and the like and displays audibly outputs and transmits the
6. image capturing device visible camera or the like 0025 The infrared sensor 10 includes for example an infrared camera or the like 0026 The laser sensor 11 measures a laser beam The laser sensor 11 includes for example a laser camera 0027 The devices 121 to 127 include for example physi cal sensors such as a thermometer hygrometer illuminom eter and electric power meter and other devices The devices 121 to 12 acquire a temperature humidity illuminance electric power information weather information schedule information and the like 0028 The BAS 13 controls monitors and manages air conditioning heat sources illuminations reception and transformation of electric energy disaster prevention secu rity and the like in a building 0029 The EMS 14 manages an environment of the energy demand prediction area 0030 The devices 251 to 25m include devices as control targets such as air conditioner lighting devices blind driving devices curtain driving devices and the like which are installed within the energy demand prediction area 0031 The input control unit 2 controls the storage device 4 to store image data 16 infrared image data infrared mea surement data 17 and laser image data laser measurement data 18 which are respectively acquired by the image sensor 9 infrared sensor 10 and laser sensor 11 0032 Also the input control unit 2 controls the storage device 4 to store device data
7. prediction data 23 and various other data 0049 The devices 251 to 25m operate based on the control data 27 The devices 251 to 25m include for example air conditioners lighting devices blind driving devices and the like 0050 FIG 2 is a block diagram showing an example of an energy demand prediction system including the energy demand prediction apparatus 1 according to this embodi ment 0051 Image sensors 91 and 92 are installed respectively for energy demand prediction areas 281 and 282 The image sensors 91 and 92 are installed on for example a ceiling of an office or an outdoor and capture images of the office The image sensors 91 and 92 may include a visible camera infra red camera and the like Image data 16 acquired by the image sensors 91 and 92 are stored in a memory area which is prepared in advance The image analysis unit 7 of an image processing server 29 analyzes the captured image data 16 to generate the human information 22a and environment infor mation 226 The image processing server 29 transmits the human information 22a and environment information 226 to the BAS 13 Note that the functions of the image analysis unit 7 may be included in the image sensors 91 and 92 In this case the need for the image processing server 29 can be obviated 0052 The BAS 13 executes generating the prediction data 23 by the prediction unit 8 and device controlling by the device control unit 24 using the human information 2
8. tion 22a according to the first embodiment more practically 0084 As described above the human information 22a includes information such as the presence absence of the person or the number of persons the distribution the amount of activity the amount of clothing the attribute name gen der body type body height age etc the position standing position seated position etc and the activity state office work transfer conversation etc of the person in the energy demand prediction area The human information 22a is acquired by analyzing the image data 16 of the image sensor 9 installed in an office The human information generation unit 7a extracts a motion ofa person by analyzing a change in luminance in time and a spatial direction of the image data 16 The human information generation unit 7a identifies a person and another object and identifies its action and behavior A target to be identified by the human information generation unit 7a is stored in a database and a learning technique is applied to the identification by the human information gen eration unit 7a 0085 The human information generation unit 7a can cal culate and estimate as the human information 22a a mea surement value at a certain point a measurement value within a designated range a value of a whole room a value of a whole floor and a value of a whole building 0086 FIG 4 is a block diagram showing an example of a human model generation uni
9. 20 2012 personal specific information of a person who is in the energy demand prediction area and the like 0037 The environment information generation unit 7b extracts a feature amount from the image data 16 infrared image data 17 and laser image data 18 and executes recog nition processing and the like based on the extracted feature amount and set criteria thus generating the environment information 225 0038 The environment information 226 includes light information such as an illuminance amount of solar radia tion a blind opening closing amount incident state of sun light and the like layout information such as the presence absence location and number of office devices the location and number of doorway and a window of an office and a location of a path the locations and numbers of heat sources and power consuming devices weather information and the like 0039 The prediction unit 8 executes the energy demand prediction based on the analysis data 22 the device data 191 to 19 the BAS data 20 the EMS data 21 and an energy demand prediction model prediction formula stored in the storage device 4 and generates prediction data 23 Then the prediction unit 8 stores the prediction data 23 in the storage device 4 In this way using the analysis data 22 which changes dynamically in the energy demand prediction a flex ible and accurate prediction can be attained 0040 The energy demand prediction model is gene
10. 2a and environment information 22b in addition to building man agement 0053 For example the prediction unit 8 executes the energy demand prediction based on the human information 22a and environment information 22b the device data 191 to 19n from other devices 121 to 12 and the BAS data 20 which is used by the BAS 13 and includes device use states electric power usages and the like 0054 Note that the building management such as OK NG determination of a human action and the energy demand prediction by the prediction unit 8 may be executed by another computer in place of the BAS 13 0055 The device control unit 24 specifies individuals who are staying in the energy demand prediction areas 281 and 282 based on the human information 22a Also the device control unit 24 specifies states sensitive to heat sensitive to the cold etc and actions during a desk work during stand talking during walking etc of individuals based on the human information 22a Furthermore the device control unit 24 executes device control which matches attributes states actions and favors of respective individuals based on Sep 20 2012 attribute information and favor information of the individuals set in the control setting data 26 0056 Thecontrol data 27 obtained as a result of the device control is transmitted to for example the control target devices 251 to 25m such as air conditioners lighting devices blind driving devices a
11. 2a may be used in combi nation with information from a building central monitoring system entry leave management system and security sys tem 0093 Furthermore the human information 22a can improve the measurement accuracy using the environment information 226 0094 The human analysis unit 35 according to this embodiment may predict a condition of another floor from a condition of one floor in association with the human infor mation 22a Fourth Embodiment 0095 This embodiment will explain modifications of the first to third embodiments described above 0096 FIG 5 isa block diagram showing an example of an arrangement of an energy demand prediction apparatus according to this embodiment FIG 5 mainly shows only components which are not shown in FIG 1 above 0097 The processor 3 executes the program 6 which is not shown in FIG 5 thereby functioning as a model selection unit 38 coefficient correction unit 39 updating unit 40 data replenishment unit 41 and singularity determination unit 42 These units will be described below 0098 Model Selection Unit 38 0099 The storage device 4 stores a plurality of energy demand prediction models 431 to 434 and model feature data 441 to 44k which respectively indicate features of the plural ity of energy demand prediction models 431 to 434 0100 The plurality of energy demand prediction models 431 to 43k and the corresponding model feature data 441 to 44k are associat
12. ATUS AND METHOD CROSS REFERENCE TO RELATED APPLICATIONS 0001 This application is a Continuation Application of PCT Application No PCT JP2011 078684 filed Dec 12 2011 and based upon and claiming the benefit of priority from prior Japanese Patent Application No 2011 057137 filed Mar 15 2011 the entire contents of which are incorporated herein by reference BACKGROUND OF THE INVENTION 0002 1 Field of the Invention 0003 An embodiment of the present invention relate to an apparatus and method predicting an energy demand of vari ous facilities 0004 2 Description of the Related Art 0005 In recent years prevention of global warming and reduction of environmental loads have received attention For example it is required to make use of capabilities of appara tuses for energy saving creation and storage equipped in facility such as a building factory and plant so as to attain unwasted and efficient energy management 0006 In order to attain energy management such as demand control an accurate energy demand prediction is required 0007 As an example of the energy demand prediction a method of executing the energy demand prediction based on information including electric power usages weather tem perature schedule business day no business day or singu larity and the like is available 0008 This conventional energy demand prediction uses for example measurement values of physical sensors suc
13. US 20120239213A1 as United States a2 Patent Application Publication o Pub No US 2012 0239213 A1 Nagata et al 43 Pub Date Sep 20 2012 54 ENERGY DEMAND PREDICTION 30 Foreign Application Priority Data APPARATUS AND METHOD Mar 15 2011 JP EE 2011 057137 76 Inventors Kazumi Nagata Fuchu shi JP Publication Classification Kenji Baba Kodaira shi JP 51 Int Cl Takaaki Enohara Hino shi JP GO6F 1 26 2006 01 Yasuo Takagi Chigasaki shi JP GD WSO anaa Kabag e a Aa an PA 700 291 Nobutaka Nishimura Koganei shi JP Shuhei Noda Fuchu shi JP 57 ABSTRACT An energy demand prediction apparatus according to an 21 Appl No 13 361 641 embodiment includes an image analysis unit and a prediction unit The image analysis unit generates analysis data includ f ing at least one of human information and environment infor 22 Filed Jan 30 2012 mation of a prediction target area based on image data acquired by an image sensor The prediction unit generates Related U S Application Data payee oai by e a energy eer DEE ased on the analysis data and an energy demand prediction 63 Continuation of application No PCT JP2011 078684 model generated using previous data corresponding to the filed on Dec 12 2011 16 Image data Seneca 27 Control data Blind driving analysis data 22a Human information 22b Environment information US 2012 0239213 Al Sep 20 2012 Sheet 1 of 4 Patent Ap
14. a case in which all values are measured using sensors 0123 Singularity Determination Unit 42 0124 The singularity determination unit 42 executes for example comparison processing with a threshold based on the acquired data such as the human information 22a envi ronment information 224 and device data 191 to 19 thus determining or predicting a singularity Then the singularity determination unit 42 stores singularity determination data 46 indicating a determination result in the storage device 4 0125 For example various components such as the model selection unit 38 coefficient correction unit 39 updating unit 40 and data replenishment unit 41 execute their processes using the singularity determination data 46 0126 A day singularity such as an anniversary of foun dation of a company which is different from a normal day has an energy demand tendency different from the normal day Hence energy demands have to be carefully predicted 0127 For example different energy demand prediction models are prepared for a business day and a singularity and it is important to recognize the singularity As a setting of the singularity a schedule may be manually input However the same energy demand prediction as that of the singularity is often exhibited on an unexpected day For this reason the singularity determination unit 42 determines or predicts a singularity based on the image data 16 acquired by the image sensor 9 Thu
15. analysis Alternatively one ora plurality of acquired data may be selected and used in image analysis 0060 A practical example will be explained below In this embodiment the human information 22a is generated mainly using the image sensor 9 and correspondence processing to a demand response for example an energy consumption reduction request from a power company and demand pre diction are executed based on this human information 22a By contrast for example the human information 22a may be generated using a thermo sensor laser sensor 11 a sensor detecting a presence of person based on a technology and the human information 22a may be used in the correspondence processing to the demand response and demand prediction 0061 In the aforementioned embodiment The energy demand is predicted based on the human information 22a and environment information 22b generated based on the image data 16 in addition to the device data 191 to 19 acquired by the device sensors 121 to 127 Thus the prediction accuracy can be improved 0062 Also using the image sensor 9 and image analysis unit 7 the number of pieces of effective information used in the energy demand prediction can be increased without installing sensors of various other types and information which changes dynamically can be effectively used in the prediction thus attaining a cost reduction 0063 In this embodiment the device control which matches an attribute of a recogn
16. ast one of head count information distribution information activity amount information clothing amount information attribute informa tion and action information of persons and Sep 20 2012 the energy demand prediction apparatus further comprises a human model generation unit that generates a human model which represents a feature of a person who are staying in the prediction target area based on the human information 6 The energy demand prediction apparatus of claim 1 further comprising a storage unit that stores a plurality of energy demand prediction models and a plurality of model feature data respectively indicating features of the plurality of energy demand prediction models in association with each other and a selection unit that selects based on the analysis data the energy demand prediction model associated with the model feature data which matches the analysis data wherein the prediction unit executes the energy demand prediction based on the analysis data and the selected energy demand prediction model 7 The energy demand prediction apparatus of claim 1 further comprising a coefficient correction unit that corrects a coefficient of the energy demand prediction model based on the analysis data 8 The energy demand prediction apparatus of claim 1 further comprising an updating unit that sequentially updates the energy demand prediction model based on the analysis data every time a predetermined p
17. device data 191 to 19 every time a predetermined period elapses For example the updating unit 40 sequentially updates the energy demand prediction models 431 to 43k using data acquired a day ahead data acquired an hour earlier and data acquired a minute earlier Updating of the energy demand prediction models 431 to 43 can use various model automatic generation techniques 0112 In this embodiment the energy demand prediction models 431 to 43 are sequentially updated using not only the human information 22a and environment information 22b acquired by the image sensor but also the device data 191 to 19n such as electric power information weather information or the like thus improving the prediction accuracy 0113 Data Replenishment Unit 41 0114 Thedata replenishment unit 41 calculates replenish ment data 45 effective for the energy demand prediction based on the acquired data such as the human information 22a environment information 22b and device data 191 to 19n Then the data replenishment unit 41 stores the replen ishment data 45 in the storage unit 4 0115 For example various components such as the model selection unit 38 coefficient correction unit 39 and updating unit 40 execute their processes using the replenishment data 45 calculated by the data replenishment unit 41 0116 For example when the electric power usages of the devices 251 to 25m cannot be directly acquired the data Sep 20 2012 replenish
18. e on the image data 16 caused by a change in illuminance and image sensor parameters Learning or updating of illuminance calculations are made using illuminance and luminance levels stored in the database As for a blind the environment information gen eration unit 7b recognizes a location of the blind by means of object recognition for the image data 16 or manual inputs Then the environment information generation unit 7b detects a change of the blind in the image data 16 and recognizes a opening closing amount and opening closing angle of the blind The state opening closing amount and opening clos ing angle of the blind calculated from the image data 16 are stored in a database Learning or updating associated with Sep 20 2012 recognition of the opening closing amount and opening clos ing angle of the blind is done using the blind state opening closing amount and opening closing angle of the image data 16 which are stored in the database 0071 The presence absence and location of an office device the number of office devices and office layout infor mation are obtained by executing recognition processing of an object such as office device and the like to the image data 16 The representative office device desk chair display PC printer partition whiteboard etc is recognized using a mea surement of an object shape a relationship of an object lay out or a learning technique A direction and a size of the object
19. ed with each other 0101 The model feature data 441 to 444 are used as cri teria required to select an appropriate model from the plural ity of energy demand prediction models 431 to 434 0102 The model selection unit 38 selects based on data such as current latest human information 22a current envi ronment information 224 and current device data 191 to 197 model feature data which matches a content indicated by the current data from the model feature data 441 to 44k Then the model selection unit 38 selects an energy demand prediction model corresponding to the selected model feature data 0103 The prediction unit 8 executes an energy demand prediction based on the energy demand prediction model selected by the model selection unit 38 US 2012 0239213 Al 0104 In this embodiment a plurality of energy demand prediction models are prepared in accordance with previous data tendencies 0105 In this embodiment an energy demand prediction model optimal to the current energy demand prediction area can be selected from the plurality of energy demand predic tion models 431 to 43k For example even when previous data which were referred to upon building up a certain energy demand prediction model include the same weather or sea son as the current weather or season when the number of persons in a room upon building up the model is different from the current number of persons in the room the model selection unit 38 gives prio
20. eriod elapses 9 The energy demand prediction apparatus of claim 1 further comprising a data replenishment unit that estimates a value used in the energy demand prediction based on the analysis data 10 The energy demand prediction apparatus of claim 1 further comprising a singularity determination unit configured to determine a singularity based on the analysis data and to generate singularity determination data indicating a determina tion result 11 An energy demand prediction method by a computer comprising generating analysis data including at least one of human information and environment information of a predic tion target area based on image data acquired by an image sensor and storing the analysis data in a storage device and generating prediction data by executing an energy demand prediction based on the analysis data stored in the stor age device and an energy demand prediction model gen erated using previous data corresponding to the analysis data and storing the prediction data in the storage device
21. formation and the like can be acquired when they are manually input by a user However using the image sensor 9 these pieces of information can be acquired in real time thus obviating the need of user s manual inputs 0073 The environment information generation unit 7b can calculate and estimate as the environment information 22b a measurement value at a certain point a measurement value within a designated range a value of a whole room a value of a whole floor and a value of a whole building 0074 FIG 3 is a block diagram showing an example of an environment model generation unit and environment analysis unit according to this embodiment Note that FIG 3 mainly shows components which are not shown in FIG 1 above 0075 The processor 3 executes the program 6 which is not shown in FIG 3 thereby implementing functions as an environment model generation unit 30 and environment analysis unit 31 0076 The environment model generation unit 30 gener ates an environment model 32 of an energy demand predic tion area using a model automatic generation technique for example a model automatic generation tool based on the environment information 22 stored in the storage device 4 For example the environment model 32 represents a feature and characteristic of the environment Then the environment model generation unit 30 stores the environment model 32 in the storage device 4 US 2012 0239213 Al 0077 The environment analys
22. h as a thermometer hygrometer illuminometer and electric power data 0009 However when the energy demand prediction is executed using only the measurement values of the physical sensors the accuracy does not often suffice BRIEF SUMMARY OF THE INVENTION Technical Problem 0010 The embodiment of the present invention have as an object to provide an energy demand prediction apparatus and method which improve an accuracy of an energy demand prediction Solution to Problem 0011 In an embodiment an energy demand prediction apparatus includes an image analysis unit and a prediction unit The image analysis unit generates analysis data includ ing at least one of human information and environment infor mation of a prediction target area based on image data acquired by an image sensor The prediction unit generates prediction data by executing an energy demand prediction Sep 20 2012 based on the analysis data and an energy demand prediction model generated using previous data corresponding to the analysis data BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING 0012 FIG 1 is a block diagram showing an example of an arrangement of an energy demand prediction apparatus according to a first embodiment 0013 FIG 2 is a block diagram showing an example of an energy demand prediction system including the energy demand prediction apparatus according to the first embodi ment 0014 FIG 3 is a block diagram showing
23. in the image data 16 change depending on a positional relationship between the image sensor 9 and the object The environment information generation unit 7b absorbs such a change and correctly recognizes the object using various learning techniques and the like Thus at the time of for example a layout change of an office or the like the need for the user to manually input information can be obviated and the environment information generation unit 7b can instanta neously and automatically recognize a new layout The image sensor 9 can be installed outdoors The environment informa tion generation unit 7b can generate weather information spatial information of a building to be analyzed and layout information of a surrounding building by analyzing the image data 16 of the outdoor Also the environment information generation unit 7b can generate information such as a longi tude and latitude of a measurement place and a direction of a building to be measured based on a positional relationship with the sun or stars 0072 Various kinds of information included in the envi ronment information 22b may be able to be acquired from various dedicated sensors However by acquiring various kinds of information by analyzing the image data 16 obtained by the image sensor 9 since the need for installing individual dedicated sensors can be obviated a cost reduction can be achieved Of the environment information 22b the layout information weather in
24. is unit 31 executes restora tion ofa three dimensional space estimation ofa temperature and humidity estimation of heat and wind air conditioning simulation and the like based on the environment model 32 and generates environment analysis data 33 Then the envi ronment analysis unit 31 stores the environment analysis data 33 in the storage device 4 0078 The output control unit 5 outputs the environment analysis data 33 to the output device 15 which is not shown in FIG 3 0079 In this embodiment the environment model 32 is built up based on the environment information 224 and res toration of the three dimensional space estimation of the temperature and humidity estimation of heat and wind air conditioning simulation and the like can be executed based on the environment model 32 0080 The environment information 22b may be used in combination with various kinds of information for example indoor and outdoor temperature and humidity values a wind speed a CO concentration weather etc acquired by nor mal sensors and information from the BAS 13 0081 Furthermore the environment information 22b can improve the measurement accuracy using the human infor mation 22a 0082 The environment analysis unit 31 of this embodi ment may predict a condition of another floor from a condi tion of one floor based on the environment information 225 Third Embodiment 0083 This embodiment will explain the human informa
25. ized individual and reflects a personal favor can be executed based on the device data 191 to 19 human information 22a environment information 22b and control setting data 26 US 2012 0239213 Al 0064 In this embodiment accurate and plenitude data that is the analysis data 22 device data 191 to 19 BAS data 20 and EMS data 21 can be used as the previous data and actual state data 0065 In this embodiment the accurate energy demand prediction model can be built up using information such as a temperature and humidity which are measured using physi cal sensors such as the device sensors 121 to 127 electric power information such as electric power usages of the devices 251 to 25m in the energy demand prediction area and the human information 22a and environment information 22b acquired based on the image sensor 9 0066 In this embodiment the energy demand prediction which can maintain an optimal energy balance can be executed at year month day hour or second intervals or in real time 0067 In this embodiment the energy demand prediction can be executed for various energy demand prediction areas 281 and 282 such as a building floors areas and zones 0068 In this embodiment the electric power information can be measured or acquired by the BAS 13 or electric power meter Furthermore in this embodiment electric power use states of the respective devices can be estimated for the respective energy demand p
26. ment unit 41 estimates the electric power usages of the devices 251 to 25m based on electric power information the human information 22a the environment information 22b and the like which can be acquired 0117 For example the data replenishment unit 41 deter mines that a personal computer which is placed in front of a working person is ON and estimates an electric power usage based on his or her attending time As for a printer and a copying machine their electric power usages are estimated in the same manner as that personal computer 0118 For example the data replenishment unit 41 esti mates an electric power user amount of each room based onan electric power user amount of a whole floor and attending conditions of respective rooms 0119 For example the data replenishment unit 41 esti mates or predicts use conditions of office devices based on a layout of each room and a distribution of the number of persons and can reflect them to the energy demand predic tion 0120 For example the data replenishment unit 41 esti mates or calculates a measurement item for an area without any sensors based on data of an area installed with a sensor 0121 For example the number of persons of an area with out any image sensor can be calculated by estimating flows of persons based on human information acquired from sur rounding image sensors 0122 Thus the prediction accuracy can be improved and cost can be reduced compared to
27. nd the like 0057 In the aforementioned energy demand prediction apparatus 1 an energy demand prediction model is generated to be able to predict an energy demand based on the human information 22a and environment information 226 in addi tion to information including the electric power usages weather temperature schedule business day no business day or singularity and the like 0058 The energy demand prediction model is built up using statistical predictions based on previous data deriva tions of regression expressions for the previous data use of physics formulas based on theories and the like For example the prediction based on the energy demand prediction model is accurately done in real time The energy demand prediction model built up with reference to the previous data executes future energy demand predictions based on actual state data of the energy demand prediction areas 281 and 282 0059 Note that in this embodiment image analysis uses data acquired by the image sensor 9 infrared sensor 10 and laser sensor 11 However all of these data need not always be used That is arbitrary one of the image sensor 9 infrared sensor 10 and laser sensor 11 may be installed Alternatively two or more sensors for at least one type of the image sensor 9 infrared sensor 10 and laser sensor 11 may be installed Alternatively data acquired by another sensor such as a thermo sensor heat source sensor may be used in image
28. plication Publication Dld Oo ww 12 ez oer 2 d YA 9 ae ee E EE Ee UOEWIOJUI USLUUOIIAUS Dep Deum EE yun yun yun uonesauad uoyesousb DUU Wun uondIpald uogeUWIOJUI Une UO ndino JusWUOJIAU3 L L b L N DL 6 SWS sva Souen AA uz aous b JOSUBS JOSE Josuas paseu josues few Sep ndino US 2012 0239213 Al Sep 20 2012 Sheet 2 of 4 Patent Application Publication yun sishjeue afew janas Buisseaoid afpu ez O Uz Iuoupuo JI BJep Giuo 17 Bugubim eyep joquos UOIEUNOJUL JUOWUONAU Zz UOHEWJOLUI uewny egz 4 yun DU Ouer 9 an sva b UOROIPSld Patent Application Publication Sep 20 2012 Sheet 3 of 4 US 2012 0239213 Al Processor Environment model RH generation unit L Environment Environment Environment information model analysis data 22b 32 33 FIG 3 Processor Human analysis a A Human Human information model 22a 36 FIG 4 Human model 34 US 2012 0239213 Al Sep 20 2012 Sheet 4 of 4 Patent Application Publication ejep ames spol ley Let Byep aimed ejep BIER uo1gUlWJalap zuawysiuajd Membe yun UODEUIU get yan fun Buyepd Up uoljoajas AueinBuig wawuswadau TTT Aen EEN Op Jossecold US 2012 0239213 Al ENERGY DEMAND PREDICTION APPAR
29. rated based on previous data including previous analysis data 22 previous device data 191 to 19 previous BAS data 20 and previous EMS data 21 and a previous energy consumption amount corresponding to this previous data Using the energy demand prediction model a future energy electric power demand prediction can be executed 0041 The device control unit 24 executes control process ing for the control target devices 251 to 25m associated with the energy demand prediction area based on the prediction data 23 the human information 22a and environment infor mation 22b in the analysis data 22 the device data 191 to 197 and control setting data 26 stored in the storage device 4 and generates control data 27 including at least one of control instructions and control values for the devices 251 to 25m Then the device control unit 24 stores the control data 27 in the storage device 4 0042 In this case the control setting data 26 includes individual user information individual attribute data and individual comfort state information of persons Also the control setting data 26 includes control values corresponding to human action states an action sensitive to heat that sen sitive to the cold during a desk work during stand talking during walking 0043 For example the device control unit 24 generates the control data 27 based on the prediction data 23 so that an energy demand fall within a predetermined value range Thus
30. rediction areas 281 and 282 based onat least one of the human information 22a and environment information 22b In response to a demand response the device control unit 24 selects a device which is not in use but whose power supply is ON based on the image data 16 from the image sensor 9 and can turn off the power supply of the selected device Therefore in this embodiment the device control unit 24 can flexibly executes control for the demand response Second Embodiment 0069 This embodiment will explain the environment information 226 according to the first embodiment more practically 0070 As described above the environment information 226 includes light information such as an illuminance amount of solar radiation blind opening closing amounts and incident amount of sunlight layout information such as the presence absence locations and number of office devices the numbers and locations of doorways and win dows and a location of a path the locations and numbers of heat sources and power consuming devices weather informa tion and the like The environment information 22b can be acquired by analyzing the image data 16 of the image sensor 9 installed in an office For example the illuminance can be calculated by setting in advance a luminance distribution on the image data 16 of a certain object under given conditions Luminance levels according to illuminance levels are stored in a database based on a change in luminanc
31. rity to a head count condition over the previous weather or season and selects an energy demand prediction model which matches the head count condition 0106 Thus the prediction accuracy can be improved 0107 Coefficient Correction Unit 39 0108 The coefficient correction unit 39 automatically cor rects coefficients of the energy demand prediction models 431 to 43k based on data such as the acquired human infor mation 22a environment information 22b and device data 191 to 19 Thus the prediction accuracy can be improved 0109 More specifically the coefficient correction unit 39 corrects the coefficients of the energy demand prediction models 431 to 43 based on the current information or infor mation including time serial changes such as an increase or decrease in the number of persons in a building or floor a weather or cloud condition and incident sunlight For example when the number of persons on a floor is increased immediately the coefficient correction unit 39 adjusts a coef ficient corresponding to the number of persons so as to build up an energy demand prediction model which matches the actual state thereby minimizing a difference between an actual energy demand and a predicted value 0110 Updating Unit 40 0111 The updating unit 40 updates sequentially updates the energy demand prediction models 431 to 43k based on the current data such as the human information 22a environment information 225 and
32. s an unexpected energy demand variation can be coped with in advance 0128 The singularity determination unit 42 determines or predicts a singularity using an increase or decrease in the number of persons a flow of persons the number of transfer ring persons a head count distribution on respective floors their time serial changes and the like included in the human information 22a mainly obtained from the image data 16 Thus energy demand tendency of a whole building whole floor and whole room can be predicted thus improving the energy demand prediction accuracy of the whole building US 2012 0239213 Al 0129 For example for a singularity an energy demand prediction model can be switched 0130 Additional advantages and modifications will readily occur to those skilled in the art Therefore the inven tion in its broader aspects is not limited to the specific details and representative embodiments shown and described herein Accordingly various modifications may be made without departing from the spirit or scope of the general inventive concept as defined by the appended claims and their equiva lents What is claimed is 1 An energy demand prediction apparatus comprising an image analysis unit that generates analysis data includ ing at least one of human information and environment information of a prediction target area based on image data acquired by an image sensor and a prediction unit that generates prediction
33. t and human analysis unit according to this embodiment Note that FIG 4 mainly shows components which are not shown in FIG 1 above 0087 The processor 3 executes the program 6 which is not shown in FIG 4 thereby implementing functions as a human model generation unit 34 and human analysis unit 35 0088 The human model generation unit 34 generates a human model 36 in an energy demand prediction area using a model automatic generation technique based on the human Sep 20 2012 information 22a stored in the storage device 4 For example the human model 36 represents a feature and characteristic of aperson Then the human model generation unit 34 stores the human model 36 in the storage device 4 0089 The human analysis unit 35 executes an action pre diction of a person and air conditioning and lighting simula tions according to the action of the person based on the human model 36 and generates human analysis data 37 Then the human analysis unit 35 stores the human analysis data 37 in the storage device 4 0090 The output control unit 5 outputs the human analysis data 37 to the output device 15 0091 In this embodiment the human model 36 in the energy demand prediction area is built up based on the human information 22a and the action prediction the air condition ing and lighting simulations according to the action of the person and the like can be executed based on the human model 36 0092 The human information 2

1Y1`

Contents

Download Pdf Manuals

Related Search

Related Contents