JP2018038206A - Attribute estimation device, attribute estimation method, and attribute estimation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an attribute estimation device, an attribute estimation method, and an attribute estimation program capable of estimating attributes of a power demand facility in a wider range.SOLUTION: An attribute estimation device comprises: an acquisition unit for acquiring monitor data including first power reception data indicating a measurement result of power consumption in each of a plurality of monitor facilities and attribute data of each monitor facility, and second power reception data indicating a measurement result of power consumption in one or more estimation target facilities; a classification unit for performing classification processing on the basis of the monitor data acquired by the acquisition unit; and an estimation unit for creating the attribute data of the estimation target facility on the basis of a result of the classification processing performed by the classification unit and the second power reception data.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an attribute estimation device, an attribute estimation method, and an attribute estimation program.

  In facilities such as general homes and offices, a device management system that manages a plurality of electrical devices by communicating with various electrical devices has been developed (for example, “ECHONET Lite standard Ver1.01”, Echonet Consortium, 2012). Issued on March 5 (Non-Patent Document 1)).

"ECHONET Lite Standard Ver 1.10", Echonet Consortium, issued on May 31, 2013

  HEMS (Home Energy Management System) and BEMS (Building Energy Management System), which are examples of the device management system described in Non-Patent Document 1, monitor the amount of power consumed in a power demand facility, and the equipment or facilities in the power demand facility By controlling the operating state of the power plant, the power consumption in the power demand facility is suppressed.

  In addition, the reform of the electric power system will completely liberalize the retailing of electric power, allowing consumers to freely select electric power retailers such as new electric power companies. As a result, the new electric power company can grasp the amount of power received at the power demand facility of the contract partner, so that a new service based on the analysis of large-scale data including the received data can be considered.

  For example, if the new power company can obtain the attributes of the power demand facility of the contract partner, it is possible to enhance the contents of such a new service or to appropriately determine the destination of the new service. However, it is generally difficult to obtain information inside the facility such as the attributes of the power demand facility from outside the facility, and the attributes of a wider range of power demand facilities are estimated from the attributes of some power demand facilities. Technology is required.

  This invention was made in order to solve the above-mentioned subject, The objective provides the attribute estimation apparatus, the attribute estimation method, and the attribute estimation program which can estimate the attribute of a more extensive power demand facility That is.

  (1) In order to solve the above-described problem, an attribute estimation apparatus according to an aspect of the present invention includes first power reception data indicating measurement results of power consumption in each of a plurality of monitor facilities, and attribute data of each of the monitor facilities. An acquisition unit that acquires monitor data including the second power reception data indicating a measurement result of power consumption in one or a plurality of estimation target facilities, and a classification process based on the monitor data acquired by the acquisition unit A classification unit to perform, and an estimation unit that creates attribute data of the estimation target facility based on a result of the classification process performed by the classification unit and the second power reception data.

  (8) In order to solve the above-described problem, an attribute estimation method according to an aspect of the present invention is an attribute estimation method in an attribute estimation apparatus, and is a first result showing a measurement result of power consumption in each of a plurality of monitor facilities. The monitor data including the received power data and the attribute data of each of the monitor facilities, and the second received data indicating the measurement result of the power consumption in the one or more estimation target facilities; and the acquired monitor data Performing a classification process based on the result, and creating attribute data of the estimation target facility based on the result of the classification process and the second power reception data.

  (9) In order to solve the above-mentioned problem, an attribute estimation program according to an aspect of the present invention is an attribute estimation program used in an attribute estimation apparatus, and measures power consumption in each of a plurality of monitor facilities. An acquisition unit that acquires first power reception data indicating results and monitor data including attribute data of each of the monitor facilities, and second power reception data indicating measurement results of power consumption in one or a plurality of estimation target facilities; A classification unit that performs classification processing based on the monitor data acquired by the acquisition unit, a result of the classification processing performed by the classification unit, and an attribute of the estimation target facility based on the second power reception data It is a program for functioning as an estimation unit that creates data.

  The present invention can be realized not only as an attribute estimation apparatus including such a characteristic processing unit but also as an attribute estimation system including an attribute estimation apparatus. Further, the present invention can be realized as a semiconductor integrated circuit that realizes part or all of the attribute estimation apparatus.

  According to the present invention, it is possible to estimate a wider range of power demand facility attributes.

FIG. 1 is a diagram showing a configuration of an attribute estimation system according to the first embodiment of the present invention. FIG. 2 is a diagram showing an example of attribute data used in the attribute estimation system according to the first embodiment of the present invention. FIG. 3 is a diagram showing a configuration of the attribute estimation apparatus in the attribute estimation system according to the first embodiment of the present invention. FIG. 4 is a diagram illustrating an example of power reception data used in the attribute estimation system according to the first embodiment of the present invention. FIG. 5 is a diagram showing an example of power reception data used in the attribute estimation system according to the first embodiment of the present invention. FIG. 6 is a diagram illustrating an example of a classification result obtained by the attribute estimation apparatus according to the first embodiment of the present invention. FIG. 7 is a diagram illustrating an example of a classification result obtained by the attribute estimation apparatus according to the first embodiment of the present invention. FIG. 8 is a diagram illustrating an example of a classification result obtained by the attribute estimation apparatus according to the first embodiment of the present invention. FIG. 9 is a diagram illustrating an example of a classification result obtained by the attribute estimation apparatus according to the first embodiment of the present invention. FIG. 10 is a flowchart that defines an operation procedure when the attribute estimation device according to the first embodiment of the present invention estimates the attributes of the estimation target facility. FIG. 11 is a diagram showing an example of power reception data used in the attribute estimation system according to the second embodiment of the present invention. FIG. 12 is a diagram showing an example of power reception data used in the attribute estimation system according to the second embodiment of the present invention. FIG. 13 is a diagram schematically showing a classification process of power reception data and attribute data in the attribute estimation apparatus according to the second embodiment of the present invention. FIG. 14 is a diagram illustrating an example of the classification result of the received power data by the attribute estimation device according to the second embodiment of the present invention. FIG. 15 is a diagram illustrating an example of the classification result of the received power data by the attribute estimation device according to the second embodiment of the present invention. FIG. 16 is a flowchart that defines an operation procedure when the attribute estimation device according to the second embodiment of the present invention estimates an attribute of a facility to be estimated.

  First, the contents of the embodiment of the present invention will be listed and described.

  (1) An attribute estimation apparatus according to an embodiment of the present invention includes monitor data including first power reception data indicating measurement results of power consumption in each of a plurality of monitor facilities, and attribute data of each of the monitor facilities; Alternatively, an acquisition unit that acquires second power reception data indicating measurement results of power consumption in a plurality of estimation target facilities, a classification unit that performs a classification process based on the monitor data acquired by the acquisition unit, and the classification An estimation unit that creates attribute data of the estimation target facility based on the result of the classification process performed by the unit and the second power reception data.

  With such a configuration, for example, the new electric power company uses the data received from all the power demanding facilities and the attribute data of the monitor facility that is a part of the all power demanding facilities to be used for billing the power charges. Since the attribute can be estimated, it is possible to obtain the attributes of all the power demanding facilities of the contracted party. Therefore, it is possible to estimate attributes of a wider range of power demand facilities. Thereby, the new electric power company can enhance the content of the new service or appropriately determine the destination of the new service.

  (2) Preferably, the classification unit receives power classification data obtained by classifying the first power reception data into a plurality of categories based on the monitor data acquired by the acquisition unit, and the attribute of each monitor facility Creating attribute classification data obtained by classifying data into a plurality of categories, and relationship data indicating a relationship between the power reception classification data and the attribute classification data, and the estimation unit includes the power reception classification data, the attribute classification data, and the relationship The attribute data of the estimation target facility is created based on the data and the second power reception data.

  With such a configuration, for example, even when the number of dimensions of the power reception data is large, the attribute data of the estimation target facility can be created more accurately.

  (3) Preferably, the first power reception data and the second power reception data indicate time-series measurement results.

  As described above, since the configuration using each power reception data reflecting the activity pattern in each power demand facility, each first power reception data can be more appropriately classified, so that the attribute data of the estimation target facility can be more accurately classified. Can be created.

  (4) More preferably, the first power reception data and the second power reception data indicate the measurement result by a smart meter.

  With such a configuration, for example, a new electric power company can use the measurement result by the smart meter for billing the electric power charge. Therefore, the attribute of the estimation target facility can be set without constructing a new acquisition route of the received power data. Can be estimated.

  (5) Preferably, the said attribute data contains the attribute regarding the apparatus of a plant | facility.

  Thus, the new power company can realize a useful new service for the device, for example, by the configuration for estimating the attribute related to the device in the estimation target facility.

  (6) Preferably, the attribute data includes an attribute relating to a resident of the facility.

  Thus, for example, the new electric power company can realize a new service suitable for the resident by estimating the attribute regarding the occupant in the estimation target facility.

  (7) Preferably, the said attribute estimation apparatus is further provided with the data preparation part which produces the recommendation data used for the recommendation regarding the electric power in the said estimation object facility based on the said attribute data produced by the said estimation part.

  With such a configuration, for example, a new electric power company can realize a better new service using recommended data.

  (8) An attribute estimation method according to an embodiment of the present invention is an attribute estimation method in an attribute estimation apparatus, and includes first received power data indicating measurement results of power consumption in each of a plurality of monitor facilities, and each of the monitors Acquiring monitor data including facility attribute data and second power receiving data indicating a measurement result of power consumption in one or a plurality of estimation target facilities, and performing a classification process based on the acquired monitor data And creating attribute data of the estimation target facility based on the result of the classification process and the second power reception data.

  With such a configuration, for example, the new electric power company uses the data received from all the power demanding facilities and the attribute data of the monitor facility that is a part of the all power demanding facilities to be used for billing the power charges. Since the attribute can be estimated, it is possible to obtain the attributes of all the power demanding facilities of the contracted party. Therefore, it is possible to estimate attributes of a wider range of power demand facilities. Thereby, the new electric power company can enhance the content of the new service or appropriately determine the destination of the new service.

  (9) An attribute estimation program according to an embodiment of the present invention is an attribute estimation program used in an attribute estimation device, and the computer receives a first power reception indicating a measurement result of power consumption in each of a plurality of monitor facilities. An acquisition unit that acquires monitor data including data and attribute data of each of the monitor facilities, and second power reception data indicating a measurement result of power consumption in one or a plurality of estimation target facilities; and acquired by the acquisition unit A classification unit that performs classification processing based on the monitor data; an estimation unit that creates attribute data of the estimation target facility based on the result of the classification processing performed by the classification unit and the second power reception data; It is a program to make it function as.

  With such a configuration, for example, the new electric power company uses the data received from all the power demanding facilities and the attribute data of the monitor facility that is a part of the all power demanding facilities to be used for billing the power charges. Since the attribute can be estimated, it is possible to obtain the attributes of all the power demanding facilities of the contracted party. Therefore, it is possible to estimate attributes of a wider range of power demand facilities. Thereby, the new electric power company can enhance the content of the new service or appropriately determine the destination of the new service.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated. Moreover, you may combine arbitrarily at least one part of embodiment described below.

<First Embodiment>
[Configuration and basic operation]
FIG. 1 is a diagram showing a configuration of an attribute estimation system according to the first embodiment of the present invention.

  With reference to FIG. 1, the attribute estimation system 301 includes an attribute estimation device 101 and a storage device 171.

  The storage device 171 is managed by, for example, a new power company that supplies power to the plurality of power demand facilities 111.

  For example, a device group 121 and a smart meter 151 are provided in the power demand facility 111. A facility ID is assigned to the power demand facility 111. In this example, facility IDs U0001 to U1000 are assigned to 1000 power demand facilities 111.

  The device group 121 includes a water heater, a cooking appliance, a photovoltaic power generation device (Photovoltaic: PV), a household storage battery, a fuel cell, an EV (Electric Vehicle), a PHEV (Plug-in Hybrid Electric Vehicle), and the like.

  More specifically, the water heater is a gas water heater, a kerosene water heater, an electric water heater, a gas engine water heater, or the like. The cooking utensils are an IH heater, a gas stove, a radiant heater, and the like.

  The smart meter 151 has, for example, a unique measuring instrument ID, measures the power consumption amount and the generated power amount in the corresponding device group 121, and sends a measurement packet including the measurement result, the measurement time, and its own measuring instrument ID to the power line. The data is transmitted to the storage device 171 by communication or the like.

  The storage device 171 stores power reception data of 1000 power demanding facilities 111 that receive power supply from a new power company. This power reception data is used, for example, for calculating a power charge and estimating a profile to be described later.

  More specifically, when the storage device 171 receives a measurement packet from the smart meter 151, the storage device 171 acquires a measurement result, a measurement time, and a measuring instrument ID from the received measurement packet.

  Further, the storage device 171 holds correspondence information indicating the correspondence relationship between the measuring instrument ID and the facility ID, and corresponds the measurement result, the measurement time, and the corresponding facility ID to the measuring instrument ID based on the correspondence information. Add and accumulate. Note that the storage device 171 may store the reception time of the measurement packet as the measurement time.

  FIG. 2 is a diagram showing an example of attribute data used in the attribute estimation system according to the first embodiment of the present invention.

  Referring to FIG. 2, the storage device 171 includes, for example, some of the power demand facilities 111 to be monitored (hereinafter also referred to as monitor facilities) out of 1000 power demand facilities 111 that receive power supply from a new power company. )) Is accumulated.

  In this example, the storage device 171 stores attribute data ADm for each of 100 monitor facilities having a facility ID of U0001 to U0100 out of 1000 power demand facilities 111.

  The attribute data ADm is information based on, for example, a questionnaire result for the manager or resident of the monitor facility, and is registered by the user.

  Specifically, the attribute data ADm includes, for example, an attribute related to the device group 121 of the power demand facility 111 and an attribute related to a resident of the power demand facility 111.

  More specifically, the attribute data ADm includes data on the number of households, a household configuration indicating age, sex, occupation, parent-child relationship, etc., and pets, as an example of attributes relating to residents of the power demand facility 111.

  Further, the attribute data ADm is, as an example of an attribute relating to the equipment of the power demand facility 111, data on a water heater, a cooking appliance, and an EV that are consumer equipment, a PV that is a power generator, and a household storage battery that is a power storage equipment. including.

  The attribute data ADm includes, as an example of an attribute related to the structure of the power demand facility 111, a house structure indicating whether the house is owned or leased, and whether it is a set or a detached house, and a house structure indicating a structure such as a wooden structure, RC, or steel frame. And data on contract power. The attribute data ADm may include building age, floor plan, and communication contract.

  FIG. 3 is a diagram showing a configuration of the attribute estimation apparatus in the attribute estimation system according to the first embodiment of the present invention.

  With reference to FIG. 3, the attribute estimation apparatus 101 includes an acquisition unit 11, an output unit 14, and a data processing unit 15. The data processing unit 15 includes a classification unit 12, an estimation unit 13, and a data creation unit 16.

  The acquisition unit 11 receives power reception data PDm indicating measurement results of power consumption in each of a plurality of monitor facilities and monitor data including attribute data ADm of each monitor facility, and power receptions indicating measurement results of power consumption in a plurality of estimation target facilities. Data PDt is acquired.

  More specifically, for example, the acquisition unit 11 acquires each attribute data ADm from the storage device 171 when a predetermined condition is satisfied. Here, the predetermined condition is when the acquisition timing of a predetermined cycle arrives, when the attribute data ADm is added, deleted, or updated.

  FIG. 4 is a diagram illustrating an example of power reception data used in the attribute estimation system according to the first embodiment of the present invention.

  Referring to FIG. 4, acquisition unit 11 acquires a facility ID of a monitor facility from each attribute data ADm, and acquires a measurement result by smart meter 151 corresponding to each acquired facility ID from storage device 171.

  Based on the acquired measurement results, the acquisition unit 11 totals the difference between the power consumption and the power generation at each measurement timing for a predetermined period, for example, a specified day, for each monitor facility. And the acquisition part 11 calculates the time average about the difference in each measurement timing, ie, average electric energy, peak electric power, dispersion | distribution, and a median for every monitor facility as the receiving data PDm. The acquisition unit 11 may calculate another statistical value regarding the difference at each measurement timing as the power reception data PDm.

  FIG. 5 is a diagram showing an example of power reception data used in the attribute estimation system according to the first embodiment of the present invention.

  Referring to FIG. 5, for example, the acquisition unit 11 is also referred to as a power demand facility 111 (hereinafter also referred to as an estimation target facility) that is different from the monitor facility among 1000 power demand facilities 111 that receive power supply from a new power company. That is, the measurement result by the smart meter 151 corresponding to the facility IDs of 900 estimation target facilities having facility IDs U0101 to U1000 is acquired from the storage device 171.

  Based on the acquired measurement results, the acquisition unit 11 aggregates the difference between the power consumption amount and the generated power amount at each measurement timing in the predetermined period for each estimation target facility. And the acquisition part 11 calculates the time average about the difference in each measurement timing, ie, average electric energy, peak power, dispersion | distribution, and a median for every estimation object facility as the receiving data PDt.

  The acquisition unit 11 creates monitor data including each attribute data ADm and each power reception data PDm, and outputs the created monitor data and each power reception data PDt to the data processing unit 15.

  The classification unit 12 in the data processing unit 15 performs the classification process based on the monitor data acquired by the acquisition unit 11.

  More specifically, for example, the classification unit 12 creates power reception attribute data in which each attribute data ADm and each power reception data PDm are collected for each monitor facility, and uses deep learning to generate a plurality of power reception attribute data. Extract features.

  For example, the classification unit 12 calculates a similarity between power reception attribute data based on a plurality of extracted feature amounts, in this case, a cosine similarity, and uses each calculated similarity to The categories are classified into the number designated by the user (hereinafter also referred to as the designated number).

  In this example, the classification unit 12 classifies each power reception attribute data into three categories C1 to C3. Note that the number of categories classified by the classification unit 12 can be changed.

  For example, the classification unit 12 extracts the characteristics of the power receiving attribute data classified into the C1 to C3 categories, and creates the power receiving attribute classification data A1 to A3 including the extracted characteristics.

  More specifically, the classifying unit 12 calculates, for example, an average value for each component as a feature for each categorized power receiving attribute data.

  Specifically, for each power receiving attribute data classified into the category of C1, the classification unit 12 averages the average power amount, the average value of the peak power, the average value of the variance, the average value of the median, An average value, an average value of age, and the like are calculated as characteristics of each component. The classification unit 12 creates power reception attribute classification data A1 including the calculated characteristics of each component.

  The classification unit 12 processes the power receiving attribute data classified into the categories C2 and C3 in the same manner, and creates power receiving attribute classification data A2 and A3.

  FIG. 6 is a diagram illustrating an example of a classification result obtained by the attribute estimation apparatus according to the first embodiment of the present invention. In FIG. 6, the vertical axis represents the average value ratio.

  FIG. 6 shows an example of average power profiling. More specifically, “C1” in FIG. 6 represents an average value obtained by subtracting the average value MA of the average power amounts of all the monitoring facilities from the average value M1 of the average power amount of each power receiving attribute data classified into the category of C1. The value divided by the value MA, that is, the average value ratio is shown. Here, the average value M1 is a feature of the average power amount included in the power receiving attribute classification data A1. The average value MA is an average value of the average power amount included in each power reception data PDm.

  Similarly, “C2” and “C3” in FIG. 6 indicate average value ratios obtained by dividing the characteristics of the average power amount included in the power receiving attribute classification data A2 and A3 by the average value of the average power amount of all the monitor facilities, respectively. .

  In this example, the characteristic of the average power amount of the category of C1 is that the average power amount of all the monitor facilities is approximately 1.5 times the standard when the average value of the average power amount of all the monitor facilities is used as a reference.

  7 to 9 are diagrams illustrating examples of classification results obtained by the attribute estimation apparatus according to the first embodiment of the present invention. 7 to 9, the vertical axis represents the average value ratio.

  FIG. 7 shows an example of profiling of PV, a household storage battery, and EV.

  FIG. 8 shows an example of pet profiling.

  FIG. 9 shows an example of contract power profiling.

  Like these, the characteristics about PV, a household storage battery, EV, a pet, and contract electric power can be recognized by the ratio with respect to the average of all the monitoring facilities.

  The classification unit 12 outputs the generated power reception attribute classification data A1 to A3 to the estimation unit 13.

  The estimation unit 13 creates attribute data ADt of the estimation target facility based on the result of the classification process performed by the classification unit 12 and each power reception data PDt.

  Specifically, the estimation unit 13 creates the attribute data ADt of the estimation target facility based on the power reception attribute classification data A1 to A3 received from the classification unit 12 and each power reception data PDt, for example, using deep learning.

  More specifically, for example, the estimation unit 13 sequentially selects the facility ID of the estimation target facility, and creates attribute data ADt for the estimation target facility (hereinafter also referred to as a selection estimation target facility) of the selected facility ID. To do.

  Specifically, the estimation unit 13 determines, for example, any one of C1 to C3 as a category for classifying the power reception data PDt of the selected estimation target facility.

  Specifically, the estimation unit 13 determines which of the power reception attribute classification data A1 to A3 the power reception data PDt of the selected estimation target facility is similar to the average power amount, peak power, variance, and median characteristics.

  More specifically, the estimation unit 13 includes the average power amount, peak power, variance, and median characteristics included in the received power attribute classification data, and the average power amount, peak power, variance, and Based on the respective similarities with the median, the category to which the power receiving data PDt of the selected estimation target facility should be classified is determined.

  For example, the estimation unit 13 extracts a component of the same type as the component of the attribute data ADm from the components included in the received power attribute classification data corresponding to the determined category, and creates attribute data ADt including the extracted component.

  The estimation unit 13 similarly creates attribute data ADt for other estimation target facilities. The estimation unit 13 outputs the created attribute data ADt for each estimation target facility to the data creation unit 16.

  Referring again to FIG. 3, the data creation unit 16 creates recommendation data used for recommendations regarding power in each estimation target facility based on, for example, each attribute data ADt created by the estimation unit 13.

  Specifically, the data creation unit 16 recommends data for the estimation target facility that indicates an optimal contract power proposal, device recommendation, PV sales, and the like based on the attribute data ADt of the target estimation target facility. And the created recommended data is output to the output unit 14.

  The output unit 14 displays the recommended data received from the data creation unit 16 on a display device (not shown) or transmits it to a wireless terminal device such as a smart phone.

[Modification of Attribute Estimation Device 101]
In the modification of the attribute estimation apparatus 101, the classification unit 12 performs a classification process using, for example, a k-means method.

  More specifically, the classification unit 12 uses the k-means method to classify each created power reception attribute data into a specified number of categories, for example, three categories C1 to C3, and to classify the power reception attribute classification data A1 to A3. create. In this k-means method, features are extracted and classified simultaneously. The classification unit 12 outputs the power receiving attribute classification data A1 to A3 to the estimation unit 13.

  The estimation unit 13 creates the attribute data ADt of the estimation target facility based on the power reception attribute classification data A1 to A3 and each power reception data PDt using the k-means method.

[Operation]
The attribute estimation apparatus 101 includes a computer, and an arithmetic processing unit such as a CPU in the computer reads and executes a program including a part or all of each step of the flowchart shown below from a memory (not shown). The program of this apparatus can be installed from the outside. The program of this device is distributed in a state stored in a recording medium.

  FIG. 10 is a flowchart that defines an operation procedure when the attribute estimation device according to the first embodiment of the present invention estimates the attributes of the estimation target facility.

  Referring to FIG. 10, attribute estimation apparatus 101 waits until a predetermined condition is satisfied (NO in step S2). When the predetermined condition is satisfied (YES in step S2), it accumulates monitor data and each received power data PDt. Obtained from the device 171 (step S4).

  Next, the attribute estimation apparatus 101 creates power reception attribute data in which the acquired attribute data ADm and each power reception data PDm are collected for each monitor facility (step S6).

  Next, the attribute estimation apparatus 101 classifies the power receiving attribute data into a specified number of categories using deep learning or the k-means method, and creates power receiving attribute classification data based on the classification result (step S8).

  Next, the attribute estimation apparatus 101 creates attribute data ADt for each estimation target facility based on the power reception attribute classification data and each power reception data PDt (step S10).

  Next, the attribute estimation apparatus 101 creates recommended data for each estimation target facility based on each created attribute data ADt, and outputs the created recommended data (step S12).

  Next, the attribute estimation apparatus 101 waits until a predetermined condition is satisfied (NO in step S2).

  In addition, although the attribute estimation apparatus according to the first embodiment of the present invention is configured to use the measurement result by the smart meter 151 as the power reception data, the present invention is not limited to this. The attribute estimation apparatus 101 may be configured to use a measurement result by HEMS or BEMS as power reception data.

  In the attribute estimation system according to the first embodiment of the present invention, a plurality of estimation target facilities are provided. However, the present invention is not limited to this. The attribute estimation system 301 may have a configuration in which one estimation target facility is provided.

  In the attribute estimation apparatus according to the first embodiment of the present invention, the classification unit 12 is configured to perform the classification process based on the monitor data. However, the present invention is not limited to this. The classification unit 12 may be configured to perform a classification process based on the monitor data and each received power data PDt.

  In the attribute estimation apparatus according to the first embodiment of the present invention, the classification unit 12 is configured to use the deep learning and the k-means method. However, the present invention is not limited to this. The classification unit 12 includes a neural network, an SVM, a decision tree, a k-NN group average method, a single connection method, a complete connection method, a Ward method, a centroid method median method, a spectral clustering, a mixed distribution model, a dimension aggregation method, and a Fuzzy c- The structure using means etc. may be sufficient.

  In the attribute estimation apparatus according to the first embodiment of the present invention, the estimation unit 13 is configured to use the deep learning and the k-means method. However, the present invention is not limited to this. The estimation unit 13 uses a neural network, correlation analysis, factor analysis, multivariate analysis, regression analysis, multiple regression analysis, principal component analysis, correspondent analysis, Painsian network, self-organization map, higher-order self-organization map, and the like. It may be a configuration.

  By the way, HEMS and BEMS which are an example of the equipment management system of a nonpatent literature 1 monitor power consumption in a power demand facility, and manage the operation state of the apparatus or equipment in a power demand facility. Reduce power consumption at demand facilities.

  In addition, the reform of the electric power system will completely liberalize the retailing of electric power, allowing consumers to freely select electric power retailers such as new electric power companies. As a result, the new electric power company can grasp the amount of power received at the power demand facility of the contract partner, so that a new service based on the analysis of large-scale data including the received data can be considered.

  For example, if the new power company can obtain the attributes of the power demand facility of the contract partner, it is possible to enhance the contents of such a new service or to appropriately determine the destination of the new service. However, it is generally difficult to obtain information inside the facility such as the attributes of the power demand facility from outside the facility, and the attributes of a wider range of power demand facilities are estimated from the attributes of some power demand facilities. Technology is required.

  On the other hand, in the attribute estimation device according to the first embodiment of the present invention, the acquisition unit 11 receives the power reception data PDm indicating the measurement result of the power consumption in each of the plurality of monitor facilities and the attribute data of each monitor facility. Monitor data including ADm and power reception data PDt indicating measurement results of power consumption in one or a plurality of estimation target facilities are acquired. The classification unit 12 performs a classification process based on the monitor data acquired by the acquisition unit 11. Then, the estimation unit 13 creates attribute data ADt of the estimation target facility based on the result of the classification process performed by the classification unit 12 and the power reception data PDt.

  With such a configuration, for example, the new electric power company uses the received power data PDm and PDt from the total power demand facility 111 and the attribute data ADm of the monitor facility that is a part of the total power demand facility 111 to be used for billing the power rate. Since the attribute of the estimation target facility can be estimated from the above, the attributes of all the power demand facilities 111 of the contracted party can be obtained. Therefore, it is possible to estimate attributes of a wider range of power demand facilities. Thereby, the new electric power company can enhance the content of the new service or appropriately determine the destination of the new service.

  Moreover, in the attribute estimation apparatus according to the first embodiment of the present invention, the power reception data PDm and the power reception data PDt indicate the measurement results by the smart meter 151.

  With such a configuration, for example, the new electric power company can use the measurement result by the smart meter 151 for billing the electric power charge, so the attribute of the estimation target facility can be obtained without constructing a new acquisition route of the received power data. Can be estimated.

  Moreover, in the attribute estimation device according to the first embodiment of the present invention, the attribute data ADm includes attributes relating to the equipment of the power demand facility 111.

  Thus, the new power company can realize a useful new service for the device, for example, by the configuration for estimating the attribute related to the device in the estimation target facility.

  Moreover, in the attribute estimation apparatus according to the first embodiment of the present invention, the attribute data ADm includes an attribute relating to the occupant of the power demand facility 111.

  Thus, for example, the new electric power company can realize a new service suitable for the resident by estimating the attribute regarding the occupant in the estimation target facility.

  Further, in the attribute estimation device according to the first embodiment of the present invention, the data creation unit 16 uses the recommended data used for the recommendation regarding the power in the estimation target facility based on the attribute data ADt created by the estimation unit 13. Create more.

  With such a configuration, for example, a new electric power company can realize a better new service using recommended data.

  Next, another embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated.

<Second Embodiment>
The present embodiment relates to an attribute estimation apparatus that uses time-series received data as compared to the attribute estimation apparatus according to the first embodiment. The contents other than those described below are the same as those of the attribute estimation apparatus according to the first embodiment.

  FIG. 11 is a diagram showing an example of power reception data used in the attribute estimation system according to the second embodiment of the present invention. FIG. 11 shows received power data PDm2 indicating the time-series measurement results of power consumption in each monitor facility.

  Referring to FIG. 11, acquisition unit 11 acquires attribute data ADm from storage device 171 when, for example, a predetermined condition is satisfied.

  The acquisition unit 11 acquires the facility ID of the monitor facility from each attribute data ADm, and acquires the measurement result by the smart meter 151 corresponding to each acquired facility ID from the storage device 171.

  Based on the acquired measurement result, the acquisition unit 11 receives the power amount of the difference between the power consumption amount of the monitor facility and the generated power amount and the power amount every 30 minutes in a specified period, for example, the specified one day. PDm2 is summarized for each monitor facility.

  FIG. 12 is a diagram showing an example of power reception data used in the attribute estimation system according to the second embodiment of the present invention. FIG. 12 shows received power data PDt2 indicating the time-series measurement result of power consumption in each estimation target facility.

  With reference to FIG. 12, the acquisition unit 11 acquires, for example, a measurement result by the smart meter 151 corresponding to the facility ID of each estimation target facility from the storage device 171.

  Based on the acquired measurement result, the acquisition unit 11 is the estimation target as the power reception data PDt2 which is the difference between the power consumption amount and the power generation amount in the estimation target facility, and the power amount every 30 minutes in the predetermined period. Summarize by facility.

  The acquisition unit 11 creates monitor data including each attribute data ADm and each power reception data PDm2, and outputs the created monitor data and each power reception data PDt2 to the data processing unit 15.

  FIG. 13 is a diagram schematically showing a classification process of power reception data and attribute data in the attribute estimation apparatus according to the second embodiment of the present invention.

  Referring to FIG. 13, the classification unit 12 includes power reception classification data obtained by classifying the power reception data PDm2 and PDt2 into a plurality of categories based on the monitor data acquired by the acquisition unit 11 and the power reception data PDt2, and each monitor facility The attribute data ADm is classified into a plurality of categories, and the relationship data RD indicating the relationship between the power receiving classification data and the attribute classification data is created.

  FIG. 14 is a diagram illustrating an example of the classification result of the received power data by the attribute estimation device according to the second embodiment of the present invention. In FIG. 14, the vertical axis represents the average amount of power received, and the horizontal axis represents time. Moreover, the time change CA of the average power reception amount indicates the average of each monitor facility.

  FIG. 15 is a diagram illustrating an example of the classification result of the received power data by the attribute estimation device according to the second embodiment of the present invention. In FIG. 15, the vertical axis represents the usage probability of electricity, and the horizontal axis represents time.

  With reference to FIGS. 13-15, the classification | category part 12 classify | categorizes each receiving data PDm2 and PDt2 into several receiving power classification data, for example using a k-means method.

  In this example, the classification unit 12 classifies each power reception data PDm2 and PDt2 into power reception classification data PDca to PDcc respectively corresponding to the categories C1 to C3. Hereinafter, the power receiving classification data PDca to PDcc are collectively referred to as a power receiving classification data group PDc.

  A feature of the power receiving classification data PDca classified as C1 is that the probability of using electricity is high at midnight. The characteristics of the power receiving classification data PDcb classified as C2 indicate that the use probability of electricity is high in the daytime. A feature of the power receiving classification data PDcc classified as C3 is that the probability of using electricity is high from evening to midnight and in the morning.

  The classification unit 12 classifies each attribute data ADm into attribute classification data ADca to ADcm using, for example, the k-means method. Hereinafter, the attribute classification data ADca to ADcm are collectively referred to as an attribute classification data group ADc.

  Further, the classification unit 12 creates relation data RD indicating the degree of relation between each of the power receiving classification data PDca to PDcc and each of the attribute classification data ADma to ADcm, for example, using the k-means method.

  The classification unit 12 outputs the generated power reception classification data group PDc, attribute classification data group ADc, and relation data RD to the estimation unit 13.

  The estimation unit 13 creates the attribute data ADt of the estimation target facility based on the power reception classification data group PDc, the attribute classification data group ADc, the relationship data RD, and each power reception data PDt2, for example, using the k-means method.

  More specifically, for example, the estimation unit 13 sequentially selects the facility ID of the estimation target facility, and creates attribute data ADt for the estimation target facility of the selected facility ID, that is, the selected estimation target facility.

  Specifically, the estimation unit 13 determines any one of C1 to C3 as a category for classifying the power reception data PDt2 of the selected estimation target facility.

  Specifically, the estimation unit 13 determines which of the power reception classification data PDca to PDcc included in the power reception classification data group PDc is the power reception data PDt2 of the selected estimation target facility.

  For example, when estimating the attributes of the selected estimation target facility, the estimation unit 13 is attribute classification data having one or a plurality of attribute classification data having a relationship with the received power classification data of the determined category and having a relationship based on the relationship data RD. To extract. Then, the estimation unit 13 weights the degree of the corresponding relationship to each extracted attribute classification data, and then calculates attribute data ADt obtained by superimposing the attribute classification data. The attribute data ADt includes characteristics of each attribute.

  The estimation unit 13 similarly creates attribute data ADt for other estimation target facilities. The estimation unit 13 outputs the created attribute data ADt for each estimation target facility to the data creation unit 16.

[Operation]
FIG. 16 is a flowchart that defines an operation procedure when the attribute estimation device according to the second embodiment of the present invention estimates an attribute of an estimation target facility.

  Referring to FIG. 16, attribute estimation apparatus 101 waits until a predetermined condition is satisfied (NO in step S102), and when the predetermined condition is satisfied (YES in step S102), it accumulates monitor data and each received power data PDt2. Obtained from the device 171 (step S104).

  Next, the attribute estimation apparatus 101 uses the k-means method to create power reception classification data obtained by classifying each power reception data PDm2 and PDt2 into a plurality of categories (step S106).

  Next, the attribute estimation apparatus 101 creates attribute classification data in which each attribute data ADm is classified into a plurality of categories using the k-means method (step S108).

  Next, the attribute estimation apparatus 101 uses the k-means method to create relationship data RD indicating the relationship between the power reception category and the attribute category based on the respective power reception data PDm2 and PDt2 and each attribute data ADm (step) S110).

  Next, the attribute estimation apparatus 101 creates attribute data ADt of each estimation target facility based on each power reception classification data, each attribute classification data, relational data RD, and each power reception data PDt2 (step S112).

  Next, the attribute estimation apparatus 101 creates recommended data for each estimation target facility based on each created attribute data ADt, and outputs the created recommended data (step S114).

  Next, the attribute estimation apparatus 101 waits until a predetermined condition is satisfied (NO in step S102).

  Note that the order of steps S106 and S108 is not limited to the above, and the order may be changed.

  Moreover, in the attribute estimation device according to the second embodiment of the present invention, the classification unit 12 receives the received power classification data obtained by classifying the received power data PDm2 and PDt2 into a plurality of categories based on the monitor data and the received power data PDt2. Although the configuration is to be created, the present invention is not limited to this. The classification unit 12 may be configured to create power reception classification data in which each power reception data PDm2 is classified into a plurality of categories based on the monitor data.

  As described above, in the attribute estimation device according to the second embodiment of the present invention, the classification unit 12 classifies each received data PDm2 into a plurality of categories based on the monitor data acquired by the acquisition unit 11. Power reception classification data, attribute classification data obtained by classifying each attribute data ADm into a plurality of categories, and relation data RD indicating the relationship between power reception classification data and attribute classification data are created. Then, the estimation unit 13 creates attribute data of the estimation target facility based on the power reception classification data, the attribute classification data, the relation data RD, and the power reception data PDt2.

  With such a configuration, for example, even when the number of dimensions of the power reception data is large, the attribute data of the estimation target facility can be created more accurately.

  In the attribute estimation apparatus according to the second embodiment of the present invention, the power reception data PDm2 and the power reception data PDt2 indicate time-series measurement results.

  As described above, since the power reception data PDm2 and the power reception data PDt2 reflecting the activity pattern in each power demand facility 111 can be classified more appropriately, the attribute data of the estimation target facility ADt can be created more accurately.

  Since other configurations and operations are the same as those of the attribute estimation apparatus according to the first embodiment, detailed description thereof will not be repeated here.

  Note that some or all of the components and operations of the devices according to the first embodiment and the second embodiment of the present invention can be combined as appropriate.

  The above embodiment should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  The above description includes the following features.

[Appendix 1]
The first power receiving data indicating the power consumption measurement result in each of the plurality of monitor facilities, the monitor data including the attribute data of each of the monitor facilities, and the second indicating the power consumption measurement result in one or a plurality of estimation target facilities. An acquisition unit for acquiring the received data of
A classification unit that performs a classification process based on the monitor data acquired by the acquisition unit;
An estimation unit that creates attribute data of the estimation target facility based on a result of the classification process performed by the classification unit and the second power reception data;
The said classification | category part is an attribute estimation apparatus which performs the said classification process using deep learning or k-means method.

DESCRIPTION OF SYMBOLS 11 Acquisition part 12 Classification part 13 Estimation part 14 Output part 15 Data processing part 16 Data preparation part 101 Attribute estimation apparatus 111 Electric power demand facility 121 Equipment group 151 Smart meter 171 Accumulation apparatus 301 Attribute estimation system

Claims (9)

The first power receiving data indicating the power consumption measurement result in each of the plurality of monitor facilities, the monitor data including the attribute data of each of the monitor facilities, and the second indicating the power consumption measurement result in one or a plurality of estimation target facilities. An acquisition unit for acquiring the received data of
A classification unit that performs a classification process based on the monitor data acquired by the acquisition unit;
An attribute estimation apparatus comprising: an estimation unit that creates attribute data of the estimation target facility based on a result of the classification process performed by the classification unit and the second power reception data. The classification unit is configured to classify each of the first power reception data into a plurality of categories based on the monitor data acquired by the acquisition unit, and classify the attribute data of each monitor facility into a plurality of categories. Creating the classified attribute classification data and the relationship data indicating the relationship between the received power classification data and the attribute classification data;
The attribute estimation device according to claim 1, wherein the estimation unit creates the attribute data of the estimation target facility based on the power reception classification data, the attribute classification data, the relation data, and the second power reception data.   The attribute estimation apparatus according to claim 1, wherein the first power reception data and the second power reception data indicate a time-series measurement result.   The attribute estimation apparatus according to claim 3, wherein the first power reception data and the second power reception data indicate the measurement result by a smart meter.   The attribute estimation apparatus according to any one of claims 1 to 4, wherein the attribute data includes an attribute related to a facility device.   The attribute estimation apparatus according to any one of claims 1 to 5, wherein the attribute data includes an attribute related to a resident of a facility. The attribute estimation device further includes:
7. The data creation unit according to claim 1, further comprising a data creation unit that creates recommendation data used for recommendation regarding power in the estimation target facility based on the attribute data created by the estimation unit. Attribute estimation device. An attribute estimation method in an attribute estimation device,
The first power receiving data indicating the power consumption measurement result in each of the plurality of monitor facilities, the monitor data including the attribute data of each of the monitor facilities, and the second indicating the power consumption measurement result in one or a plurality of estimation target facilities. Obtaining the received data of
Performing a classification process based on the acquired monitor data;
Creating attribute data of the facility to be estimated based on a result of the classification process and the second power reception data. An attribute estimation program used in an attribute estimation device,
Computer
The first power receiving data indicating the power consumption measurement result in each of the plurality of monitor facilities, the monitor data including the attribute data of each of the monitor facilities, and the second indicating the power consumption measurement result in one or a plurality of estimation target facilities. An acquisition unit for acquiring the received data of
A classification unit that performs a classification process based on the monitor data acquired by the acquisition unit;
An estimation unit that creates attribute data of the estimation target facility based on the result of the classification process performed by the classification unit and the second power reception data;
Attribute estimation program to function as

Images