WO2017104044A1 - Dispositif, procédé et système de gestion de santé à distance - Google Patents

Dispositif, procédé et système de gestion de santé à distance Download PDF

Info

Publication number
WO2017104044A1
WO2017104044A1 PCT/JP2015/085348 JP2015085348W WO2017104044A1 WO 2017104044 A1 WO2017104044 A1 WO 2017104044A1 JP 2015085348 W JP2015085348 W JP 2015085348W WO 2017104044 A1 WO2017104044 A1 WO 2017104044A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
server
sensor
resident
power
Prior art date
Application number
PCT/JP2015/085348
Other languages
English (en)
Japanese (ja)
Inventor
晶子 松田
正明 中島
吉川　暁
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Priority to JP2017555944A priority Critical patent/JPWO2017104044A1/ja
Priority to PCT/JP2015/085348 priority patent/WO2017104044A1/fr
Publication of WO2017104044A1 publication Critical patent/WO2017104044A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/02—Alarms for ensuring the safety of persons
    • G08B21/04—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B25/00—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems
    • G08B25/01—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems characterised by the transmission medium
    • G08B25/04—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems characterised by the transmission medium using a single signalling line, e.g. in a closed loop
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M11/00—Telephonic communication systems specially adapted for combination with other electrical systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04Q—SELECTING
    • H04Q9/00—Arrangements in telecontrol or telemetry systems for selectively calling a substation from a main station, in which substation desired apparatus is selected for applying a control signal thereto or for obtaining measured values therefrom

Definitions

  • the present invention relates to a technique for managing the health of residents in remote locations.
  • Patent Document 1 describes a technique for determining that a resident is in an emergency state when the activity stoppage time or the activity continuation time is equal to or longer than a predetermined time based on the measurement data of the energy consumption of the target household.
  • Patent Document 2 describes a technique for determining a resident's abnormality when the time when the moving object detection unit is not detected reaches a predetermined time.
  • a remote health management system includes a power meter that measures power consumption in a specific power line in a house, and a resident's activity in a specific place in the house.
  • a sensor to detect, a communication device connected to the sensor, and a server connected to the communication device via a communication network.
  • the server repeatedly acquires power data based on the output of the power meter, the server registers the power data in a database, and the server repeatedly acquires sensor data based on the output of the sensor from the communication device.
  • the server registers the sensor data in the database, the server generates a plurality of pattern data based on the database, and the server generates the current pattern data from the plurality of pattern data.
  • the designated pattern data suitable for the situation of the house is selected, and the server is configured based on the sensor data, the electric power data, and the designated pattern data, from among a plurality of preset states of the resident of the house. Estimate the state.
  • the plurality of states include the resident being at home and abnormal, the resident being at home and normal, and the resident being absent from the house.
  • the configuration of the remote health management system is shown. The operation of each part in the housing system is shown.
  • a power data acquisition process is shown.
  • the sensor data acquisition process is shown.
  • a pattern generation process is shown.
  • the relationship between a pattern generation process, a pattern estimation process, and a resident state estimation process is shown. Details of the condition matrix are shown.
  • the power increase condition is shown. Indicates the power divergence condition.
  • the resident status screen is shown. An outline of the correction process is shown.
  • Fig. 1 shows the configuration of a remote health management system.
  • the remote health management system includes at least one housing system 10, at least one support system 30, and a processing system 40.
  • the housing system 10 is provided in a house.
  • the residential system 10 includes an HGW (Home Gate Way) 110, a power meter 120, a heat pump 130, a temperature sensor 140, a human sensor 150, a behavior sensor 155, a door sensor 160, a smart plug 170, and a home appliance. 180, display operation panel 210, and telephone 220.
  • HGW Home Gate Way
  • the housing system 10 may not include any of the heat pump 130, the temperature sensor 140, the human sensor 150, the behavior sensor 155, the door sensor 160, and the smart plug 170.
  • the power meter 120 measures the power consumption of the specific location by integrating the power consumption of the specific location in the house over a preset measurement time.
  • the power meter 120a measures the amount of power consumed by the house.
  • the power meter 120 b measures the power consumption of the heat pump 130.
  • the power from the distribution line of the power system is supplied to the heat pump 130 via the power meter 120a and the power meter 120b, and is supplied to the home appliance 180 via the power meter 120a and the smart plug 170.
  • the housing system 10 may not include the power meter 120b.
  • the heat pump 130 is an air conditioner, a refrigerator, an electric water heater, or the like.
  • the temperature sensor 140 measures an air temperature such as room temperature.
  • the temperature sensor 140 may be included in the heat pump 130.
  • the door sensor 160 is provided on a door at a specific place in the house and continuously detects whether or not the door is closed.
  • the door sensor 160 may be a combination of a reed switch and a magnet, or may be another switch.
  • the human sensor 150 is provided at a door of an entrance (entrance) of the house.
  • the human sensor 150 is provided at a specific place in the house and continuously detects whether or not a person is active at the place.
  • the human sensor 150 may be an infrared sensor or another sensor such as an ultrasonic sensor. Further, the human sensor 150 may detect the presence of a target or may detect a temperature.
  • the human sensor 150 includes a human sensor provided inside a door of an entrance / exit of the house (for example, an entrance hall) and a human sensor provided in a bathroom, toilet, living room, or the like.
  • the behavior sensor 155 is attached to the resident's body, and the behavior of the person is detected by an acceleration sensor such as a three-axis sensor (X, Y, Z).
  • the behavior sensor 155 may include other biological sensors (pulse, respiration, etc.). Further, the behavior sensor 155 may include a position measurement system (GPS: Global Positioning System).
  • GPS Global Positioning System
  • the household electrical appliance 180 is a device that operates by consuming electric power.
  • the home appliance 180 is an electric device that operates in response to an operation by a resident, such as an electric kettle, a microwave oven, or a TV.
  • Smart plug 170 is provided between home appliance 180 and an outlet, and measures the power consumption of home appliance 180.
  • the smart plug 170 may measure the power consumption amount by integrating the power consumption.
  • the display operation panel 210 is connected to the HGW 110 via a communication network such as a wireless LAN, communicates with the HGW 110, displays information from the HGW 110, and accepts input from a resident.
  • the display operation panel 210 may be a tablet, a smartphone, a PC, or the like. Instead of the display operation panel 210, other input devices and display devices may be used.
  • the display operation panel 210 may be connected to the communication network 60.
  • the telephone 220 is connected to the telephone line 70 and is used for a call by a resident in an emergency or the like.
  • the telephone 220 may be connected to the communication network 60 when the telephone line 70 is disconnected.
  • the HGW 110 includes a memory 111, a CPU (Central Processing Unit) 112, a communication interface 113, and a sensor interface 114.
  • the memory 111 stores programs and data.
  • the program may be stored in a computer-readable storage medium and installed in the HGW 110 from the storage medium.
  • the CPU 112 performs processing of the HGW 110 according to the program in the memory 111.
  • the communication interface 113 is connected to the communication network 60 and performs communication based on an instruction from the CPU 112.
  • the sensor interface 114 is connected to the power meter 120, the temperature sensor 140, the human sensor 150, the door sensor 160, and the smart plug 170, and communicates with these components.
  • the HGW 110 acquires the power consumption amount from each of the power meter 120 and the smart plug 170 at each preset power data measurement time, and transmits the power data indicating the acquired power consumption amount to the server 410.
  • the HGW 110 acquires measurement results and control results from the temperature sensor 140, the human sensor 150, and the door sensor 160, creates sensor data indicating the measurement results and control results, and stores the sensor data. Send to server 410.
  • the housing system 10 may include other sensors such as a pressure sensor and a bed sensor.
  • the pressure sensor is provided on a sofa in a living room, and detects pressure due to the use of a resident's sofa.
  • the bed sensor is provided on a bed in a bedroom and detects the sleep and wake-up of a resident.
  • the support system 30 is provided in the facilities of supporters for residents.
  • the supporter may be a person in charge or an assistant of the remote health management service, or may be a related person such as a relative of the resident.
  • the support system 30 includes a display / operation panel 310 and a telephone 320.
  • the display operation panel 310 communicates with the server 410 to display information received from the server 410 and accept input from a supporter.
  • the display operation panel 310 may be a tablet, a smartphone, a PC, or the like. Instead of the display operation panel 310, other input devices and display devices may be used.
  • the telephone 320 is connected to the telephone line 70 and is used for a call by a supporter in an emergency or the like.
  • the display operation panel 310 may be portable or may be provided in another place. Thereby, a supporter can know a resident's abnormality anywhere.
  • the processing system 40 includes a server 410, a telephone 420, and a storage 430.
  • the server 410 is connected to the communication network 60.
  • the storage 430 stores the database 440.
  • the telephone 420 is connected to the telephone line 70 and is used for a telephone call by an administrator of the processing system 40 in an emergency or the like.
  • the server 410 includes a memory 411, a CPU 412, and a communication interface 413.
  • the memory 411 stores programs and data.
  • the program may be stored in a computer-readable storage medium and installed on the server 410 from the storage medium.
  • the CPU 412 performs processing of the server 410 according to the program in the memory 411.
  • the communication interface 413 is connected to each unit of the processing system 40 to the communication network 60 and performs communication based on an instruction from the CPU 412.
  • the server 410 may be connected to the management terminal via the communication network 60.
  • the management terminal includes a memory, a CPU, a communication interface, an input device, and an output device.
  • the management terminal accepts an input to the input device from the administrator of the server 410 and transmits it to the server 410. Further, the management terminal displays information received from the server 410 on the display device.
  • the server 410 acquires power data from the HGW 110 of each house and registers the power data in the database 440. Further, the server 410 acquires sensor data from the HGW 110 of each house and registers the sensor data in the database 440. Furthermore, the server 410 generates pattern data indicating patterns of residents' activities and power consumption based on the database 440 and registers the pattern data in the database 440. Further, the server 410 estimates a resident's abnormality based on the power data and sensor data acquired from the HGW 110 and the database 440. The abnormality is, for example, a state where the resident is at home and cannot move.
  • processing system 40 may include another computer, and the computer may include the storage 430. Further, the processing system 40 may include another computer, and the computer may provide the estimation result by the server 410 to the support system 30 or the like via the communication network 60.
  • the supporter can check the status of the resident based on the information from the server 410 and can rush to the house or send an ambulance to the house.
  • the resident may input the resident information regarding the housing system 10 and the resident to the display operation panel 210, the HGW 110 may transmit the resident information to the server 410, and the server 410 may register the resident information in the database 440.
  • Resident information may include, for example, age, sickness, type of home appliance 180, preset pattern data, information on supporters, number of residents in a house, and the like.
  • the supporter may input the resident information using the display operation panel 310 and transmit it to the server 410.
  • the server 410 receives the resident information and registers the resident information in the database 440.
  • the manager may input the resident information to the server 410 and register the resident information in the database 440 based on the contract information with the resident.
  • the resident inputs an outing schedule indicating the time range in which the resident goes out to the display operation panel 210 using the display operation panel 210, the HGW 110 transmits the outing schedule to the server 410, and the server 410 stores the outing schedule in the database 440. You may register.
  • the remote health management system may be connected to a DR (Demand Response) system 50 via the communication network 60.
  • the DR system 50 performs DR on the housing system 10 based on the database 440 when the power supply and demand is tight.
  • the number of each part in the remote health management system may be two or more.
  • a supporter who is away from the house can know the abnormal health of the resident.
  • Fig. 2 shows the operation of each part in the housing system.
  • the horizontal axis in this figure indicates time.
  • This figure shows the sensor data transmission timing at which the HGW 110 transmits sensor data to the server 410, the output of the human sensor 150 near the door, the output of the door sensor 160, and the operation of the kettle connected to the smart plug 170.
  • the operation of the TV connected to the smart plug 170, the operation of the heat pump 130 by the resident, the output of the human sensor 150 in the living room, and the output of the human sensor 150 in the toilet are shown.
  • the sensor data transmission timing interval is a preset sensor data measurement time.
  • the HGW 110 records the operation of each device within the measurement period from the last sensor data transmission timing to the next sensor data transmission timing as a measurement period.
  • the output of the human sensor 150 is, for example, a value lower than a preset human sensor output threshold when no human movement is detected, and when a human movement is detected, a preset human sensor is detected.
  • the value is equal to or greater than the sensor output threshold.
  • the HGW 110 monitors the output of the human sensor 150, detects that the output of the human sensor 150 is equal to or higher than the human sensor output threshold (rise) as a detection start event, and the output of the human sensor 150 is human. A detection end event is detected when the output is less than the sensation sensor output threshold (falling). Further, the HGW 110 counts the number of detection start events and the number of detection end events within the measurement period, and includes the number of detection start events and the number of detection end events in the sensor data. Note that the HGW 110 may detect either the detection start event or the detection end event.
  • the output of the door sensor 160 is, for example, a value lower than a preset door sensor output threshold when the door is closed, and a value equal to or greater than a preset door sensor output threshold when the door is open.
  • the HGW 110 detects that the output of the door sensor 160 is equal to or greater than the door sensor output threshold (rise) as an open event, and detects that the output of the door sensor 160 is less than the door sensor output threshold (fall) as a close event. . Further, the HGW 110 counts the number of open events and the number of closed events within the measurement period, and includes the number of open events and the number of closed events in the sensor data. Note that the HGW 110 may detect one of an open event and a close event. The HGW 110 may include the output of the door sensor 160 at the sensor data transmission timing in the sensor data.
  • the HGW 110 may estimate the home status of the resident, such as whether the resident is at home or a change in the number of resident people in the house, based on the sensor data.
  • the HGW 110 may include the home state at the sensor data transmission timing in the sensor data. For example, in the HGW 110, an activity is detected by the human sensor 150 near the door, an opening / closing is then detected by the door sensor 160, and then an activity is detected by the human sensor 150 near the door for a preset time or more. If not, it may be estimated that the number of people at home has decreased.
  • the HGW 110 can intermittently transmit sensor data by detecting the number of events from the output, and the event Can be transmitted without omission.
  • the HGW 110 can generate sensor data by compressing sensor output data. Thereby, even when there are many houses to be managed, the load on the communication network 60 and the server 410 can be suppressed.
  • the HGW 110 repeatedly performs a power data transmission process every time a preset power data measurement time elapses.
  • the power data measurement time is, for example, 1 minute.
  • FIG. 3 shows the power data acquisition process
  • the HGW 110 acquires the output of the power meter 120 and acquires the output of the smart plug 170 (S220).
  • the HGW 110 calculates power consumption by integrating power consumption.
  • the HGW 110 creates power data indicating the acquired power consumption, and transmits the power data to the server 410 (S230).
  • the power data includes power consumption of the target house.
  • the power data may include the power consumption amount of the heat pump 130 or the power consumption amount of the home appliance 180.
  • the power meter 120 may transmit power data indicating the power consumption amount to the server 410.
  • the power meter 120 may transmit power data to the server 410 via the dedicated communication network 60.
  • the power meter 120 may transmit power data to another system such as MDMS (Meter Data Management) and the server 410 may acquire power data from the other system.
  • MDMS Method Data Management
  • the server 410 can acquire the power consumption measured at each power data measurement time and accumulate it in the database 440 as time series data.
  • FIG. 4 shows sensor data acquisition processing
  • This sequence shows the operation of the server 410 and the HGW 110.
  • the server 410 sequentially selects one house from among the houses to be managed as the target house, and performs an acquisition process for the target house.
  • the server 410 repeatedly performs the acquisition process for one house every time the sensor data measurement time elapses.
  • the server 410 may perform the acquisition process at a different timing for each house, may divide the house to be managed into a plurality of groups, and perform the acquisition process at a different timing for each group.
  • the server 410 transmits a sensor data request indicating the sensor data measurement time to the HGW 110 of the target house (S110).
  • the sensor data measurement time is, for example, 5 minutes.
  • the sensor data measurement time is set in units of minutes, and the sensor data request indicates “5”.
  • the HGW 110 starts a measurement period of the length of the sensor data measurement time (S120), and when the measurement period ends, transmits the sensor data obtained in the measurement period to the server 410 ( S130).
  • the server 410 when receiving the sensor data from the target house, registers the sensor data in the database 440 (S140).
  • the server 410 accumulates the number of open events and the number of closed events from a specific point in time, calculates the cumulative number of open events and the number of closed events, and registers them in the database 440 as sensor data. May be.
  • the server 410 determines that a failure in the housing system 10 or communication has occurred if sensor data cannot be received even after a preset reception waiting time has elapsed since the transmission of the sensor data request.
  • the reception standby time is, for example, a value obtained by adding a preset time to the sensor data measurement time.
  • the server 410 performs a pattern estimation process for selecting pattern data suitable for the current situation of the target house (S150).
  • the server 410 performs a resident state estimation process for estimating a resident state that is a resident state of the target house (S160). For example, the resident status indicates whether the user is at home or away.
  • the server 410 performs output processing for outputting the resident state obtained by the resident state estimation processing (S170), and ends this sequence.
  • the center data request may not include the sensor data measurement time.
  • the HGW 110 when receiving the sensor data request, transmits the sensor data obtained in the measurement period to the server 410 with the measurement period from the previous reception of the sensor data request to the present.
  • the server 410 can repeatedly acquire sensor data from each house and store it in the database 440. Moreover, the server 410 can detect a resident's abnormality whenever it acquires sensor data. Further, the server 410 determines the sensor data transmission timing for each house and transmits a sensor data request, so that the load on the server 410 can be distributed even when there are many houses to be managed.
  • the processing system 40 may include a plurality of servers 410.
  • different housing groups are set as management targets for a plurality of servers 410. Thereby, the load of each server 410 can be distributed.
  • the HGW 110 and the server 410 may not perform the power data acquisition process.
  • the HGW 110 transmits sensor data and power data to the server.
  • the HGW 110 accumulates power data within the measurement period.
  • the server 410 performs pattern generation processing for creating a plurality of pattern data based on the database 440 at the pattern generation timing set by the administrator.
  • the pattern generation timing may be a point in time when a predetermined time has elapsed from the start of data accumulation in the database 440, may be a preset regular time, or the server 410 may generate a pattern from the management terminal. It may be the time when a request for processing is received.
  • FIG. 5 shows pattern generation processing
  • a plurality of periods are set in advance.
  • the period is represented by a combination of season, day of the week, time zone, and the like. For example, one period is represented as 6 to 8 o'clock on Monday in winter.
  • a combination of a house and a period is called a classification.
  • the server 410 selects a target house from a plurality of managed houses, selects a target period from a plurality of periods, and selects a combination of the target house and the target period as a target classification (S310). . Thereafter, the server 410 acquires power data corresponding to the target classification from the power data in the database 440, and acquires power consumption data from the acquired power data (S320).
  • the power consumption data includes the total power consumption that is the overall power consumption of the target house, the activity power consumption that is the power consumption related to the resident's activity among the power consumption of the target house, It includes the device power consumption that is the power consumption.
  • the server 410 calculates the activity power consumption by subtracting the power consumption of the heat pump 130 from the total power consumption.
  • the server 410 analyzes the characteristics of the time series data of the activity power consumption amount, and the activity power consumption according to the analysis result. You may extract the time series data of the apparatus power consumption of each household electrical appliance 180 from the quantity time series data.
  • the time series data of the device power consumption of the household electrical appliance 180 has a unique frequency characteristic for each type.
  • the frequency characteristics of the device power consumption for each type are set in advance.
  • the server 410 calculates the frequency characteristic of the active power consumption by Fourier transform of the time series data of the active power consumption, and uses the preset frequency characteristics of the active power consumption to calculate the device consumption from the active power consumption.
  • the component of the electric energy can be extracted.
  • the amount of power consumed by the heat pump 130 is small in periodic fluctuations, such as increasing at a constant rate of change under the control of the heat pump 130 itself, and has little correlation with residents' activities. Therefore, the server 410 can obtain a power consumption amount that is highly correlated with a resident's activity by calculating the power consumption amount.
  • the equipment power consumption is extracted from the activity power consumption, so that the equipment power consumption is calculated from the total power consumption. Compared with the case of extracting the power consumption amount, the accuracy of extracting the device power consumption amount can be improved.
  • the total power consumption may be used instead of the activity power consumption.
  • the power consumption data may include the device use start time and device use end time of each home appliance 180 by the resident.
  • the server 410 detects a time when the device power consumption of the home appliance 180 is equal to or greater than a preset device power consumption threshold as the device use start time, and the device power consumption is the device power consumption threshold. The time when it becomes less than is detected as the device use end time.
  • time-series data of device power consumption for each type may be set in advance as reference device power consumption data.
  • the server 410 performs pattern matching with the reference device power consumption data for the time series data of the activity power consumption, and from the period when the obtained correlation is higher than a preset value, the device use start time and the device You may detect use end time.
  • the HGW 110 may acquire power consumption data from the output of the power meter 120 and transmit the power consumption data to the server 410.
  • the server 410 acquires power consumption data corresponding to the target classification and sensor data corresponding to the target classification.
  • the power consumption data and the sensor data are collectively referred to as measurement data.
  • the server 410 calculates pattern data corresponding to the target classification based on the acquired measurement data (S330).
  • the pattern data includes a power consumption average value that is an average of the power consumption, a power consumption range indicating a range of the power consumption, and an activity time zone.
  • the server 410 calculates the average ⁇ and standard deviation ⁇ of the power consumption at a plurality of times (samples) for the power consumption data corresponding to the target classification, and sets ⁇ as the power consumption average value. Thereafter, the server 410 calculates a power consumption range determined by the power consumption upper limit value ⁇ + n ⁇ and the power consumption lower limit value ⁇ n ⁇ using a preset coefficient n.
  • the pattern data may include ⁇ or n ⁇ instead of the power consumption range.
  • the server 410 calculates the resident's activity time zone from the power consumption data and sensor data corresponding to the target classification.
  • the activity time zone is a range of time when the resident is active in the house.
  • the server 410 uses the power consumption data and sensor data corresponding to the target classification to calculate a time range in which an activity condition described later is satisfied as an activity time zone.
  • the server 410 calculates the average and standard deviation of the temperature measured by the temperature sensor 140 among the sensor data corresponding to the target classification, and uses the temperature upper limit value and the temperature lower limit value in the same manner as the power consumption range. Calculate the fixed temperature range.
  • the temperature range may be set in advance according to the period.
  • the pattern data may include a device use start time, a device use end time, and a device use period length that is a length from the device use start time to the device use end time.
  • the pattern data may be data indicating the configuration and parameters of artificial intelligence such as a neural network.
  • the server 410 learns the neural network using the measurement data corresponding to the target classification.
  • the server 410 calculates the number of people living in the target house based on the outing schedule. It is possible to estimate and calculate pattern data for each number of people at home.
  • the server 410 may estimate the number of people at home based on measurement data corresponding to the target classification. For example, when the home appliance 180 is a rice cooker, the device usage period length of the rice cooker varies depending on the number of people at home. Therefore, the server 410 calculates the device usage period length of the rice cooker for each number of people at home.
  • the pattern data may include reference power consumption data indicating a waveform of the power consumption data of the target classification.
  • the server 410 may calculate an average waveform of power consumption data having a preset length as reference power consumption data.
  • the server 410 may determine a plurality of classifications by clustering (cluster analysis) of measurement data in the database 440. For example, the server 410 acquires measurement data from the database 440 for each period of a preset length, calculates feature data indicating the characteristics of each measurement data, and classifies the feature data into a plurality of clusters by clustering.
  • the pattern data may be calculated based on the measurement data representing each cluster. In this case, the pattern data may include feature data.
  • the feature data may be an average or standard deviation of the power consumption data, a frequency characteristic obtained by Fourier transform of the power consumption data, an occurrence frequency of each event in the sensor data, or the like.
  • the server 410 may classify the clusters into the number of home-stayers by clustering and generate pattern data of the number of home-stayers based on each cluster.
  • the server 410 determines whether pattern data has been generated for all classifications (S350). When it is determined that the pattern data has not yet been generated for all the classifications (S350: N), the server 410 returns the process to S310 and selects the next target classification. If it is determined that pattern data has been generated for all classifications (S350: Y), the server 410 ends this flow.
  • pattern generation process by classifying the measurement data, it is possible to classify the housing situation and prepare pattern data serving as a criterion for determining the resident state for each classification.
  • FIG. 6 shows the relationship among the pattern generation process, the pattern estimation process, and the resident state estimation process.
  • data based on the human sensor 150 is called human sensor data
  • data based on the door sensor 160 is called door sensor data
  • data based on the temperature sensor 140 is called temperature sensor data.
  • the server 410 acquires the latest power consumption data as target power consumption data from the database 440, and acquires the latest sensor data as target sensor data from the database 440.
  • the target power consumption data and the target sensor data are collectively referred to as target measurement data.
  • the server 410 may use the latest time-series power consumption data having a preset length as the target power consumption data, or may use the latest time-series sensor data having the length of the determination time as the target sensor data. It is good.
  • the server 410 selects a target period that is a period corresponding to the current time. Thereafter, the server 410 selects pattern data corresponding to the classification indicated by the target house and the target period from the database 440 as designated pattern data.
  • the server 410 includes an activity time zone, a power consumption average value, a power consumption amount range, and a temperature range included in the designated pattern data, respectively, as a designated activity time zone, a designated power consumption average value, a designated power consumption range, Acquired as the specified temperature range.
  • the server 410 calculates target feature data indicating the features of the target measurement data, selects a cluster corresponding to the target feature data, and corresponds to the selected cluster. Pattern data may be selected as designated pattern data.
  • the server 410 estimates the number of people at home and sets the pattern data corresponding to the number of people at home. You may select as designated pattern data.
  • the server 410 may acquire a going-out schedule corresponding to the current time from the database 440 and estimate the number of people at home.
  • reference device power consumption data for each number of people at home may be set in advance for a certain type of home appliance 180.
  • the reference device power consumption data when the number of people at home is one may be set as a basic waveform, and the reference device power consumption data for every two or more people at home may be set as an applied waveform.
  • the server 410 performs pattern matching with the reference device power consumption data for each number of people on the time series data of the activity power consumption, selects the reference device power consumption data with the highest correlation, A corresponding number of people at home may be selected.
  • the server 410 may extract the device usage period length and select the pattern data having the closest device usage period length in the same manner as the pattern generation process.
  • the server 410 can detect that the number of people at home is one in the target house where a plurality of residents live by estimating the number of people at home. As a result, the server 410 can detect an abnormality of a home person when there is no other family member.
  • the server 410 can acquire pattern data suitable for the situation of the target house. Further, when the pattern data indicates the activity time zone, it can be quickly determined whether or not the activity is normal. Further, when the pattern data indicates the power consumption range, it can be quickly determined whether or not the power consumption is normal.
  • the server 410 estimates the resident state of the target house using the condition matrix.
  • the condition matrix indicates a combination of conditions for multidimensional input.
  • the output of the condition matrix is the resident state.
  • a case where the condition 551 in the condition matrix is satisfied and the resident state 561 that is the output of the condition matrix is abnormal is shown.
  • Input includes specified pattern data, target power consumption data, and target sensor data.
  • the designated pattern data includes a designated activity time zone, a designated power consumption average value, a designated power consumption range, and a designated temperature range.
  • the target power consumption data includes the target power consumption.
  • the target power consumption here is the activity power consumption.
  • target power consumption may be total power consumption.
  • the target power consumption may be the device power consumption.
  • FIG. 7 shows a condition matrix
  • the condition matrix indicates, for example, a combination of the following multiple conditions.
  • the activity time condition which is a condition for the current time, is that the current time is within the specified activity time zone.
  • the sensor detection condition which is a condition for target sensor data, is that at least one of the number of open events, the number of closed events, the number of detection start events, and the number of detection end events is positive. Thereby, the server 410 can estimate that the resident is active.
  • the outing condition which is a condition for the target sensor data, is that the number of open events is positive, the number of closed events is positive, and the number of detection end events near the door is positive.
  • the server 410 can estimate that the resident has opened and closed the door and the resident has left the doorway.
  • the open condition which is a condition for the target sensor data, is that the cumulative open event count is one more than the cumulative closed event count. Accordingly, the server 410 can estimate that the resident has opened and closed the door and the resident has left the door.
  • the abnormal temperature condition which is a condition for the target sensor data, is that the temperature measured by the temperature sensor 140 is outside the specified temperature range. Thereby, the server 410 can estimate that the specific location of the target house is not maintained at an appropriate temperature.
  • the power usage condition which is a condition for the target power consumption data, is that the target power consumption is equal to or greater than a preset power consumption threshold.
  • the power increase condition which is a condition for the target power consumption data, is that the target power consumption is equal to or greater than the specified power consumption average value.
  • FIG. 8 shows the power increase condition.
  • the horizontal axis indicates time
  • the vertical axis indicates target power consumption.
  • the broken line indicates the specified power consumption average value.
  • the power divergence condition which is a condition for the target power consumption data, is that the target power consumption is outside the specified power consumption range.
  • FIG. 9 shows power divergence conditions.
  • the horizontal axis indicates time
  • the vertical axis indicates target power consumption.
  • the broken line indicates the average value of power consumption during the target period
  • the upper double line indicates the upper limit value of the specified power consumption range
  • the lower double line indicates the lower limit value of the specified power consumption range. Show. Thereby, the server 410 can estimate that the usage status of the power by the resident is different from the designated pattern data.
  • the activity condition is that at least one of the sensor detection condition and the power use condition is satisfied. Thereby, the server 410 can estimate that the resident is active.
  • the abnormal activity condition is that the activity time condition is not satisfied and the activity condition is satisfied.
  • the server 410 can estimate that the resident is active outside the designated activity time zone.
  • the abnormal stop condition is that the sensor detection condition is continuously satisfied over a preset stop time after the sensor detection condition is satisfied. Thereby, the server 410 can estimate that a resident does not move in a specific place.
  • the server 410 calculates a time during which the release condition is continuously satisfied as the release time based on the target sensor data. For example, the server 410 substitutes 0 for the opening time in the resident state estimation process in which the opening condition is satisfied, and if the opening condition is continuously satisfied in the subsequent resident state estimation process, the sensor data measurement is performed in the opening time. Add time.
  • the abnormal opening condition is that the opening time is not less than a preset opening time threshold. Thereby, the server 410 can estimate the opening of the door due to the abnormality of the resident or the intrusion of the suspicious person.
  • the abnormal condition is that at least one of an abnormal activity condition, a power divergence condition, an abnormal stop condition, an abnormal open condition, and an abnormal temperature condition is satisfied.
  • the server 410 calculates the time during which the activity condition is not satisfied continuously as the outing time based on the target power consumption data and the target sensor data. For example, the server 410 substitutes 0 for the outing time in the resident state estimation process in which the outing condition is satisfied, and if the activity condition is not continuously satisfied in the subsequent resident state estimation process, the sensor data measurement is performed during the outing time. Add time.
  • the absence condition is that the going-out time is not less than a preset going-out time threshold.
  • the normal activity condition is that the activity time condition is met and the activity condition is met. Thereby, the server 410 can estimate that the resident is active during the designated activity time zone.
  • the normal condition is that at least one of a normal activity condition and a power increase condition is satisfied.
  • the server 410 estimates that the resident state is abnormal.
  • the server 410 estimates that the resident state is normal.
  • the server 410 estimates that the resident state is absent.
  • priorities may be set in advance for these conditions.
  • the resident state is set as abnormal, normal, and absent in order from the highest priority.
  • the server 410 selects a resident state having a high priority.
  • the server 410 registers the resident status estimation result in the database 440.
  • the estimation result includes a resident ID, a resident state, an estimated time, and an estimation reason.
  • the estimation reason indicates, for example, a condition established in the condition matrix.
  • server 410 may output a resident state indicating inactivity as an estimation result, or may not output a resident state estimation result when none of the conditions of the resident state is satisfied.
  • condition matrix may include device usage conditions.
  • the device use condition is that the device is used during the measurement period based on the device use start time and the device use end time. When the device use condition is satisfied, the activity condition is satisfied.
  • condition matrices may be set in advance, and the pattern data may indicate the condition matrix.
  • the resident state estimation process uses the condition matrix selected by the pattern estimation process.
  • the server 410 may perform the resident state estimation process when the number of residents of the target house is 1 or when it is estimated that the number of people living at home is 1 based on the outing schedule. Further, the server 410 estimates that the number of people staying at home is 0 based on the going-out schedule, and if a normal condition or an abnormal condition is established, the server 410 estimates that an illegal intrusion or a resident has fallen and cannot go out, and the support system. An alarm is transmitted to 30 display operation panels 310.
  • the server 410 uses the neural network instead of the condition matrix, and inputs the target measurement data to the neural network to estimate the resident state. .
  • the server 410 is informed that the resident is at home and abnormal, the resident is at home and normal, the resident is absent from the house, and the resident state. Can be estimated.
  • the server 410 can reduce erroneous detection of abnormality by using the power consumption data and the sensor data.
  • the server 410 performs the estimation using the pattern data and the condition matrix, so that the process becomes simple and the resident state can be estimated quickly. Further, the server 410 can reduce erroneous detection of abnormality by switching the pattern data according to the situation.
  • the display operation panel 310 acquires the resident information and the resident state data indicating the result of the resident state estimation process from the server 410 based on the operation by the supporter, and displays the resident state screen based on the resident state data.
  • the display operation panel 210 may acquire the resident status data from the server 410 based on an operation by the resident, and display the resident status screen based on the resident status data. Further, the server 410 may transmit the resident status data to the display operation panel 310 and the display operation panel 210, and display the resident status screen on the display operation panel 310 and the display operation panel 210.
  • FIG. 10 shows the resident status screen.
  • the resident status screen includes a resident name 511, a supporter name 512, a resident status data time 513, an update button 514, an estimation status 521, a sensor status 522, a supporter list 523, a power consumption waveform 531, a device usage status 532, and sensor data. 533, a day activity button 541, an activity monitor button 542, a situation contact button 543, and an actual situation contact button 544 are included. Note that the resident status screen may indicate a determination result of a specific condition in the condition matrix.
  • the resident name 511 indicates the name and ID of the resident of the target house based on the resident information.
  • the supporter name 512 indicates the name of the supporter.
  • Resident status data time 513 indicates an analysis time that is the time of the resident status data.
  • the resident status data time 513 may be the time of the latest resident status data, or may be the time designated by the supporter.
  • the estimated state 521 indicates the resident state estimated by the resident state estimation process.
  • the estimation situation 521 may include an estimation reason.
  • the sensor status 522 indicates the status of each sensor in the target house. For example, the sensor status 522 indicates that the battery of the human sensor 150 is low and the battery needs to be replaced.
  • the supporter list 523 indicates a list of supporters who can use the display operation panel 310. For example, a supporter who can use the display operation panel 310 is a person in charge of a remote health management service (support staff).
  • the power consumption waveform 531 shows power consumption data of the target house.
  • the device usage status 532 indicates a device usage period for each home appliance 180.
  • the sensor data 533 indicates the measurement result by each sensor in the target house.
  • the display operation panel 310 acquires the resident status data for one day from the server 410, and displays the resident status screen based on the acquired resident status data.
  • the display operation panel 310 acquires the latest resident status data from the server 410 and displays a resident status screen based on the acquired resident status data.
  • the display operation panel 310 acquires the estimation result by the resident status estimation process from the server 410 and displays the estimated status 521.
  • the display operation panel 310 performs a correction process for correcting the estimation result by the resident state estimation process, and transmits the correction result to the server 410.
  • the number of people at home may be displayed on the resident state screen.
  • the server 410 transmits a notification indicating the abnormality to the supporter or the like in the output process.
  • the notification is, for example, an e-mail.
  • the processing system 40 may include a Web server that provides a service for the support system 30 and the housing system 10 to acquire resident status data.
  • the server 410 may not transmit the resident status data to the support system 30 or the housing system 10.
  • a supporter makes a phone call to each house regularly to determine whether the resident's condition is abnormal. According to the above output processing, the burden on the supporter can be reduced. Moreover, the abnormality of a resident's state can be quickly communicated to a supporter, and the possibility of dealing with the abnormality can be increased. Moreover, the supporter can confirm the cause of a resident state by displaying the information based on electric power data, the information based on sensor data, and the resident state on a resident state screen. Further, by transmitting the sensor data obtained by processing the sensor output by the HGW 110 without using a camera or the like to the server, the situation can be grasped while considering privacy.
  • the resident status is displayed on the display operation panel 210 using the actual status communication button 544 on the resident status screen. input.
  • the HGW 110 transmits a correction instruction indicating the corrected resident status to the server 410.
  • the supporter may input a correction instruction on the display operation panel 310.
  • FIG. 11 shows an outline of the correction process.
  • This figure shows the pre-correction resident state 561 in the above-described condition matrix and the post-correction resident state 562 shown in the correction instruction.
  • the server 410 that has received the correction instruction specifies the cause 551 of the pre-correction resident state 561 from the condition matrix, and corrects the data related to the cause 551 in the pattern data. For example, when the cause 551 is the establishment of the abnormal activity condition, the server 410 changes the activity time zone to a time zone including the current time. Thereby, the normal activity condition of the cause 552 is established, and the estimation result of the resident state becomes the corrected resident state 562.
  • the server 410 performs pattern generation processing using the latest measurement data having a preset length, and performs pattern generation processing using measurement data different from the previous pattern generation processing.
  • the pattern data may be updated. Thereby, the server 410 can adjust the pattern data to the latest situation.
  • the server 410 re-learns the neural network using the corrected resident state 562.
  • the resident may input a correction processing instruction to the display operation panel 210, and the display operation panel 210 may perform the correction processing.
  • the server 410 may update the pattern data by the same process as the correction process. For example, when it is determined that the resident status is abnormal, the supporter requests an ambulance, and the resident is transported by ambulance, the supporter inputs the ambulance transport to the display operation panel 310 and transmits it to the server 410. Also good. The server 410 may register the measurement data at this time as pattern data, and then notify the supporter as an emergency when the measurement data is similar to the pattern data.
  • the pattern data can be made to follow a change in the pattern of the resident's activity in the target house, a change in the home appliance 180, and the like.
  • the server 410 can improve the estimation accuracy in the subsequent resident state estimation processing.
  • the DR system 50 may perform DR using database 440.
  • the DR system 50 acquires power consumption data for each house in the target period of DR from the pattern data, and based on the acquired power consumption data, predicts a reduction in power consumption, and calculates power consumption. Select housing to reduce. Since the remote health management system and the DR system share the database 440, the costs of the two systems can be reduced.
  • Memory A sensor for detecting a resident's activity at a specific place in the house; A sensor interface connected to the sensor; A communication interface connected to the server via a communication network; A processor connected to the memory, the sensor interface, and the communication interface; With The processor detects rising and falling edges of the sensor output; The processor counts the number of rising and falling edges of the sensor output, The processor transmits sensor data including the rising number and the falling number to the server. Communication device.
  • the server and the remote health management device correspond to the server 410 and the like.
  • the sensors correspond to the human sensor 150, the behavior sensor 155, the door sensor 160, the temperature sensor 140, and the like.
  • the communication device corresponds to the HGW 110 or the like.
  • the processor corresponds to the CPU 412 and the like.
  • the terminal device corresponds to the display operation panel 210, the display operation panel 210, and the like.
  • the electric device corresponds to the home appliance 180 or the like.

Landscapes

  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Emergency Management (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Gerontology & Geriatric Medicine (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Telephonic Communication Services (AREA)
  • Alarm Systems (AREA)
  • Selective Calling Equipment (AREA)
  • Emergency Alarm Devices (AREA)

Abstract

La présente invention a pour objet d'estimer une anomalie affectant un résident, rapidement et avec une bonne précision. Un serveur: acquiert de façon répétée des données d'électricité d'après la sortie d'un compteur d'électricité; inscrit les données d'électricité dans une base de données; acquiert de façon répétée des données de capteurs d'après une sortie de capteurs provenant d'un dispositif de communication; inscrit les données de capteurs dans la base de données; génère une pluralité d'instances de données de schéma d'après la base de données; sélectionne, parmi la pluralité d'instances de données de schéma, une instance désignée de données de schéma qui correspond à la situation présente dans une résidence; et, d'après les données de capteurs, les données d'électricité et l'instance désignée de données de schéma, estime l'état d'un résident de la résidence parmi une pluralité d'états prédéfinis.
PCT/JP2015/085348 2015-12-17 2015-12-17 Dispositif, procédé et système de gestion de santé à distance WO2017104044A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
JP2017555944A JPWO2017104044A1 (ja) 2015-12-17 2015-12-17 遠隔健康管理装置、遠隔健康管理方法、および遠隔健康管理システム
PCT/JP2015/085348 WO2017104044A1 (fr) 2015-12-17 2015-12-17 Dispositif, procédé et système de gestion de santé à distance

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2015/085348 WO2017104044A1 (fr) 2015-12-17 2015-12-17 Dispositif, procédé et système de gestion de santé à distance

Publications (1)

Publication Number Publication Date
WO2017104044A1 true WO2017104044A1 (fr) 2017-06-22

Family

ID=59056194

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2015/085348 WO2017104044A1 (fr) 2015-12-17 2015-12-17 Dispositif, procédé et système de gestion de santé à distance

Country Status (2)

Country Link
JP (1) JPWO2017104044A1 (fr)
WO (1) WO2017104044A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2019153295A (ja) * 2018-03-01 2019-09-12 大阪瓦斯株式会社 自己管理支援システム
JP6830298B1 (ja) * 2020-04-30 2021-02-17 株式会社Ｊｄｓｃ 情報処理システム、情報処理装置、情報処理方法、及びプログラム
KR20220066712A (ko) * 2020-11-16 2022-05-24 주식회사 산들정보통신 빅데이터 인공지능 기반 응급상황 모니터링 시스템 및 방법

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001216585A (ja) * 2000-02-03 2001-08-10 Sekisui Chem Co Ltd 異常行動判定システム
JP2007183890A (ja) * 2006-01-10 2007-07-19 Chugoku Electric Power Co Inc:The 生活状況監視システム、装置、方法およびプログラム
JP2015109044A (ja) * 2013-12-05 2015-06-11 パナソニックＩｐマネジメント株式会社 見守り通知装置
JP2015176303A (ja) * 2014-03-14 2015-10-05 大阪瓦斯株式会社 行動判定システム、セキュリティシステム及び居住者見守りシステム

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001216585A (ja) * 2000-02-03 2001-08-10 Sekisui Chem Co Ltd 異常行動判定システム
JP2007183890A (ja) * 2006-01-10 2007-07-19 Chugoku Electric Power Co Inc:The 生活状況監視システム、装置、方法およびプログラム
JP2015109044A (ja) * 2013-12-05 2015-06-11 パナソニックＩｐマネジメント株式会社 見守り通知装置
JP2015176303A (ja) * 2014-03-14 2015-10-05 大阪瓦斯株式会社 行動判定システム、セキュリティシステム及び居住者見守りシステム

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2019153295A (ja) * 2018-03-01 2019-09-12 大阪瓦斯株式会社 自己管理支援システム
JP7216569B2 (ja) 2018-03-01 2023-02-01 大阪瓦斯株式会社 自己管理支援システム
JP6830298B1 (ja) * 2020-04-30 2021-02-17 株式会社Ｊｄｓｃ 情報処理システム、情報処理装置、情報処理方法、及びプログラム
JP2021174477A (ja) * 2020-04-30 2021-11-01 株式会社Jdsc 情報処理システム、情報処理装置、情報処理方法、及びプログラム
KR20220066712A (ko) * 2020-11-16 2022-05-24 주식회사 산들정보통신 빅데이터 인공지능 기반 응급상황 모니터링 시스템 및 방법
KR102436434B1 (ko) * 2020-11-16 2022-08-25 주식회사 산들정보통신 빅데이터 인공지능 기반 응급상황 모니터링 시스템 및 방법

Also Published As

Publication number Publication date
JPWO2017104044A1 (ja) 2018-06-21

Similar Documents

Publication Publication Date Title
JP5058504B2 (ja) 住居の中の人物の遠隔人物追跡方法とその装置
CN109843173B (zh) 用于监测人的日常生活活动的系统和方法
Alcalá et al. Detecting anomalies in activities of daily living of elderly residents via energy disaggregation and cox processes
WO2015127491A1 (fr) Système de surveillance
JP2007265017A (ja) 高齢者安否情報生成システム
US20160116512A1 (en) Method and system for monitoring energy consumption
US20160371593A1 (en) Living activity inference device, and program
JP6197258B2 (ja) 行動予測装置、プログラム
WO2017104044A1 (fr) Dispositif, procédé et système de gestion de santé à distance
JP6493828B2 (ja) 外出支援装置、およびプログラム
Patrono et al. An innovative approach for monitoring elderly behavior by detecting home appliance's usage
JP2016218801A (ja) 行動推定装置、行動推定方法、及び行動推定プログラム
EP2733647B1 (fr) Système de contrôle de puissance sur demande, programme de système de contrôle de puissance sur demande, et support d'enregistrement lisible par ordinateur enregistré avec ledit programme
JP6452593B2 (ja) 情報処理装置及び情報処理方法
JP2003281655A (ja) 家電機器利用モニタリング装置
JP6418525B2 (ja) 就寝監視装置、およびプログラム
KR102465304B1 (ko) Ai 스위치와 ai 생활 정보기를 활용한 상황 기반 인공지능 스마트 홈 시스템
JP6830298B1 (ja) 情報処理システム、情報処理装置、情報処理方法、及びプログラム
JP5920933B2 (ja) 独居者見守りシステム、独居者見守り方法及びプログラム
JP2009276847A (ja) 居住者生存情報通知システム
JPWO2019058590A1 (ja) 見守りシステム、見守り方法、冷蔵庫、及び通信端末
JP2015089163A (ja) 行動評価装置、行動評価システム、行動評価方法、プログラム
Moshtaghi et al. Monitoring personal safety by unobtrusively detecting unusual periods of inactivity
JP6947064B2 (ja) 見守り装置、見守り方法、および見守りプログラム
US20200098471A1 (en) Actions based on customer premises data

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15910734

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2017555944

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15910734

Country of ref document: EP

Kind code of ref document: A1