JP7045901B2 - Predictive control device, predictive control system and predictive control method - Google Patents

Description

The present invention relates to a predictive control device, a predictive control system and a predictive control method.

When collecting, delivering, delivering, etc. various products (including documents), the delivery company may visit your home, and in addition to checking the various conditions (including health conditions) of the inhabitants, protection is also performed in the same way as above. People may visit their homes.

It is possible for a visitor or the like to make a promise with a resident or a guardian by designating a date and time in advance, but the visit may be forced to occur suddenly. In the latter case, the number of times when the resident is not at home (when he / she is away) is overwhelmingly larger than that in the former case, and he / she cannot make contact with the target resident, resulting in a wasteful step.

This leads to a decrease in visit efficiency, and the originally necessary processing by visit (delivery, status confirmation, etc.) is delayed.

For this reason, in order to perform processing more efficiently in a situation where a sudden visit is unavoidable, a technique is disclosed that predicts the situation at home and enables the visit to be made at a time when the tendency to stay at home is high.

Japanese Unexamined Patent Publication No. 2012-181789 Japanese Unexamined Patent Publication No. 2017-62765 Japanese Patent Application Laid-Open No. 2003-288442

The above-mentioned Patent Document 1 is a technique for calculating the probability of being at home, which calculates information indicating a home status using information on the amount of power used detected by an automatic meter reading system, and is information indicating the home status. It is disclosed that the probability of staying at home is calculated in consideration of the behavioral characteristics of each household.

Further, Patent Document 2 makes it possible to predict the home and absence of a person with high accuracy, and obtains the power data of the building at a certain first time to obtain the first time. Rather, it predicts whether or not a person is in the building at the first time from the correspondence between the power data of the past building and the information indicating whether or not the person was actually in the building.

As described above, Patent Document 1 and Patent Document 2 describe predicting whether a person will be at home by using information on electric power.

In addition, Patent Document 3 calculates and predicts the date and time of staying at home, which is likely to be at home, based on the visit history data.

With such prior art, it is difficult to predict at home, especially in an environment where information on electric power cannot be obtained in Patent Document 1 and Patent Document 2, and past visit history data is required in Patent Document 3. However, it is not always possible to predict the home situation, and even if it can be predicted, the prediction result will be low in accuracy.

The present invention is a predictive control device and prediction that makes it possible to predict the home status of a facility with higher accuracy, not only for a facility that can acquire energy information, but also for a facility that cannot acquire this energy information. It is an object of the present invention to provide a control system and a predictive control method.

Forcibly, the present invention makes it possible to make a more optimal visit schedule and call schedule.

In order to achieve the above object, the predictive control device according to claim 1 of the present invention is an attribute information storage means for storing attribute information about a target person to be visited by a visitor and a facility where the target person can stay at home. The energy information to be consumed is measured by the instrument, and the acquisition means for acquiring the energy information from the instrument and the subject at the time of measuring the energy information based on the energy information acquired by the acquisition means. The attribute information corresponds to the first home status prediction means for predicting the home status, the home status predicted by the first home status prediction means, and the attribute information regarding the target person in the home status. A means for creating prediction reference data used for home prediction of another attached target person, and a means for storing the home status of the target person in the attribute information storage means when the energy information of the facility cannot be acquired. It is characterized by comprising a second home status prediction means for predicting by referring to the prediction reference data based on the attribute information of the target person.

Further, the predictive control device according to claim 2 of the present invention further includes, for the invention of claim 1, a visit history storage means for storing the visit history of the subject to the facility, and the visit history includes the visit history. Is characterized by including a case where the presence / absence of a visit to the facility of the subject is memorized and a case where the presence / absence of the visit is not memorized.

Further, in the prediction control device according to claim 3 of the present invention, with respect to the invention of claim 2, the first home situation prediction means is the visit history storage means together with the energy information of the subject. It is characterized in that the home status of the subject is predicted by using the memory of the visit history of the subject to the facility.

Further, the prediction control device according to claim 4 of the present invention cannot acquire the energy information in the facility and the visit is made by the second home situation prediction means with respect to the invention of claim 2. It is characterized in predicting the home status of the subject when the presence or absence of the visit is not stored in the visit history of the subject to the facility by the history storage means.

Further, the prediction control device according to claim 5 of the present invention is based on at least one of the first home situation prediction means and the second home situation prediction means for the invention of claim 1 or 4. Based on the attribute information of the target person who predicted the home status, the second home is further provided with a generation means for generating predictive reference data capable of predicting the home status of another target person having the attribute information. The situation prediction means is characterized in that the home status of the target person is predicted by using the prediction reference data generated by the generation means based on the attribute information of the target person for the facility visited by the visitor. And.

Further, the predictive control system according to claim 6 of the present invention includes an attribute information storage means for storing attribute information about a target person to be visited by a visitor and a home prediction means for predicting the home status of the target person. A predictive control device provided, an energy utilization device provided in a facility visited by the visitor and operating using energy, and a measuring means for measuring energy information consumed by the energy utilization device are provided. The home-based prediction means further comprises a first home-based situation predicting means for predicting the home-based situation of the target person at the time when the energy information is measured based on the energy information measured by the measuring means. The prediction control device is at home of another target person to which the attribute information is associated with the home situation predicted by the first home situation prediction means and the attribute information about the target person in the home situation. Based on the creation means for creating the prediction reference data used for prediction and the attribute information of the target person to store the home status of the target person in the attribute information storage means when the energy information cannot be measured at the facility. It is further provided with a second home prediction means for making a prediction by referring to the prediction reference data.

Further, the predictive control method according to claim 7 of the present invention stores attribute information about a target person to be visited by a visitor in a storage medium, and energy information consumed in a facility where the target person can stay at home. The step that the predictive control device measures, the step that the predictive control device predicts the home situation of the target person at the time when the energy information is measured based on the measured energy information, the predicted home situation, and A step in which the prediction control device creates prediction reference data used for home prediction of another target person to which the attribute information is associated with the attribute information related to the target person in the home situation, the energy information in the facility. When the measurement cannot be performed, the step includes a step of predicting the home status of the target person by the prediction control device with reference to the prediction reference data based on the attribute information of the target person.

According to the present invention, not only a facility that can acquire energy information but also a facility that cannot acquire this energy information can predict the home status of the facility with higher accuracy. Play.

The system block diagram which shows the predictive control system in embodiment of this invention. The figure which shows the user information stored in the user management database provided in the predictive control system in embodiment of this invention. The figure which shows the energy consumption data stored in the energy data management database provided in the predictive control system in embodiment of this invention. The figure which shows the visit history information stored in the visit history management database provided in the predictive control system in embodiment of this invention. The schematic diagram which simplifies the predictive control process performed by the predictive control apparatus in embodiment of this invention. The figure which shows the predictive reference data used for the predictive control process performed by the predictive control apparatus in embodiment of this invention. The flowchart which shows the flow of the predictive control process performed by the predictive control apparatus in embodiment of this invention.

Hereinafter, an embodiment of the predictive control device, the predictive control system, and the predictive control method according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a system configuration diagram showing a predictive control system according to an embodiment of the present invention.

In FIG. 1, the predictive control system includes a predictive control device 100, an information management server 200, an information communication control device 300, an instrument 400 with a communication function (hereinafter referred to as “instrument 400”), and an in-home system 500. The in-house system 500 shows the entire system provided in a building, a facility, or the like, and the instrument 400 is provided corresponding to the in-house system 500 and measures energy information (information about energy). Is shown.

In addition, the in-house system 500 includes individual houses (including single-family homes (also referred to as "detached houses"), apartment houses (also referred to as "aggregates")), commercial facilities such as stores, and indoors and outdoors represented by buildings. It indicates a system of facilities where humans can stay at home (hereinafter, also simply referred to as "building", "inside the house", "home", and "visited place"), which are provided regardless of the situation. More specifically, the home system 500 includes a home energy management device 501 and various energy utilization devices 502. This home energy management device 501 is also called HEMS (Home Energy Management System), is connected to the energy utilization device 502 directly or indirectly, and manages and monitors the information obtained from the energy utilization device 502. It is a control device that can communicate with the home energy management device 501. The home energy management device 501 is not an indispensable configuration requirement, but is a device that can be provided as needed.

The energy utilization device 502 is a device that can be used by residents, managers, and the like in the facility, and indicates all devices capable of communicating with the home energy management device 501. The energy utilization device 502 operates using not only devices (PCs, mobile phones, etc.) and display displays that perform various network communications, but also electric energy that is a force necessary for operating the energy utilization device 502. Devices (electrical equipment), electrical appliances (TVs, air conditioners, refrigerators, telephones, etc.), electrical equipment (solar power generation systems, secondary batteries, electric vehicles, etc.), etc. fall under this category. Of course, it is also possible to supplementally use energy such as gas and petroleum (gasoline, kerosene) together with this electric energy. Details will be described later. For this reason, the electric energy required to operate the energy utilization device 502 is called "main energy (main energy, first energy)" and is supplementarily used together with this "main energy". Energy such as oil and oil may also be referred to as "secondary energy (secondary energy)". Further, each of "main energy (first energy)" and "secondary energy (second energy)" is also collectively referred to as "energy (usage amount, operation amount, work amount)".
Applies to.

Further, in the home energy management device 501, it is possible to manage the communication status as well as the amount of energy used (consumed) by the energy utilization device 502. Information on the energy used at this time and information on the communication status (these information may be collectively referred to as "energy information") may be displayed on the display. Of course, it is also possible to display the energy information required by the home energy management device 501 itself on the display.

From the above, the home system 500 equipped with the home energy management device 501 and various energy utilization devices 502 is provided at the visit destination where the visitor can visit, and the above energy information can be collectively managed. It is a system. In addition, since the visitor visits the inhabitants and managers who live in the facility to be visited, the inhabitants and managers are referred to as "visitors", "targets", or "users (energy). User) ”.

Subsequently, the instrument 400 is composed of a measuring unit that measures (measures) the amount of energy used, a communication unit that performs communication, and the like, and receives energy information from the home energy management device 501 via the communication unit (home). By (including both the active transmission of data by the energy management device 501 and the transmission of data by the home energy management device 501 in response to a request from the instrument 400), that is, when the energy information is measured. The measuring unit is a device that measures "energy consumption". This communication unit transmits energy information regarding the continuous energy consumption to the information communication control device 300 at a predetermined timing (immediately or after a predetermined time (for example, 30 minutes) has elapsed). Since this energy information is the data used for the home determination condition having the highest priority, which will be described later, it is also referred to as "highly credible data used for predicting the home situation". As shown above, since the home energy management device 501 is not an indispensable configuration requirement, if the home energy management device 501 is not provided, the instrument 400 receives energy information directly from the energy utilization device 502. Will be done.

At this time, the instrument 400 and the information communication control device 300 are connected by either wired or wireless, and in particular, a specific frequency band (especially, via a relay device (concentrator, etc.) installed on a utility pole or the like). For example, mutual communication using radio waves in the 920 MHz band) is possible.

The instrument 400 mainly indicates an electric meter that measures the amount of electricity used for electric energy. Further, in addition to the main energy which is the electric energy, when a secondary energy is supplementarily used in the home determination performed by the prediction control device 100 described later, an instrument capable of measuring the secondary energy for each type of the secondary energy is provided. It is possible to provide.

The information communication control device 300 is connected to each instrument 400 in a loop type topology, and receives information on energy usage from each instrument 400 to communicate with the prediction control device 100 or the information management server 200. (Information transmission, information transmission / reception). The information communication control device 300 enables the prediction control device 100 or the information management server 200 to communicate with the instrument 400 regardless of the communication method (instrument 400 alone, grid configuration).

Then, the information management server 200 collects (aggregates, meter reading) the amount of energy used per predetermined time (for example, 15 minutes, 30 minutes, 1 hour) in each facility received from the instrument 400 via the information communication control device 300. ) And manage it. Further, the information management server 200 can also perform a process of calculating a cost for the amount of energy used.

A storage medium such as a user management database 201 (customer DB) and an energy data management database 202 (collected data DB) is connected to the information management server 200 at least, and a visit history management database 203 (visit history DB) described later is connected to the information management server 200. ) Storage medium can also be connected. The user management database 201 stores information about visitors (users) who are using energy (hereinafter referred to as "user information") (for example, identification information, name, address, family structure, energy contract details, etc.). Further, in the energy data management database 202, the energy usage data in each facility received from the instrument 400 is stored in chronological order in association with the user.

At this time, an example of user information (customer information) stored in the user management database 201 is shown in FIG. 2, and an example of energy usage data stored in the energy data management database 202 is shown in FIG. 3, which will be described later.

When the information management server 200 receives a request for transmission of information on the energy usage from the predictive control device 100, which will be described later, in a state where the energy usage in each facility is managed, the user ID of the user is the user management database. If it is registered in 201 and the energy usage amount for the user is managed by the energy data management database 202, the energy usage amount is returned to the predictive control device 100.

If the user ID is not registered in the user management database 201, the information management server 200 responds to the prediction control device 100 to that effect. Further, if the user ID is registered in the user management database 201 but the energy usage amount for the user is not managed in the energy data management database 202, the information management server 200 manages the user for the prediction control device 100. The user attribute information registered in the database 201 is transmitted.

The predictive control device 100 is a device capable of predicting the home situation by performing predictive control processing for all users registered in the user management database 201. Specifically, the predictive control device 100 determines whether or not the user is at home (“100%” or “0%”) for each predetermined date and time and for each time zone, or a process for calculating the probability of being at home (calculating the probability of absence). It is possible to predict the home situation by performing the processing (which may be a process to be performed) (hereinafter, also referred to simply as “home determination” or “home prediction”). That is, the home status is represented by the presence / absence of home and / or the probability of being home, and the prediction control device 100 generates a “home status list” which is a home prediction result that aggregates the home status for each user.

At least the visit history management database 203 (visit history DB) is connected to the prediction control device 100, and the user management database 201 (customer DB) and the energy data management database 202 (collection data DB) can also be connected. .. The visit history management database 203 manages the date and time of the visit and the result (at home, absent) of the visit for each user identification information (user ID) that identifies the user, and the details are shown in FIG.

The predictive control device 100 can perform predictive control processing using energy usage data and user attribute information received from the information management server 200, as well as past visit history and racks for users in the visit history management database 203. If the electric history is registered, it is also possible to perform the prediction control process using these history information.

Further, in the prediction control processing using the user attribute information, the prediction control device 100 previously obtains prediction reference data consisting of the home status for the combination of the items of the user attribute information based on the home prediction result in the prediction control processing. It is possible to perform the process of creating (aggregating). The prediction control device 100 can predict the home status of the user having the user attribute information by using the prediction reference data and the user attribute information acquired from the information management server 200.

An example of the energy consumption data at this time is shown in FIG. 3, FIG. 4 shows the past visit history, and FIG. 6 shows the prediction reference data.

It is possible to connect a terminal such as a PC or a mobile to the prediction control device 100 directly or indirectly via a network network, and receive a home prediction request (input) from the terminal. It is possible. The terminal at this time is also referred to as a "home prediction request source", and can refer to a "home status list" which is a home prediction result (output) related to the prediction control processing performed by the prediction control device 100. Of course, it is also possible to output this home prediction result to another device or a predetermined medium (paper medium, storage medium, etc.) different from the “home prediction request source”.

The predictive control process performed by the predictive control device 100 makes it possible to perform home-based prediction with high accuracy by preferentially using home-based determination conditions with high-precision home-based prediction, and also makes it possible to perform home-based determination conditions according to each user. By using, it is possible to predict the home of the specified user or all users.

Specifically, as the highest priority home judgment condition (priority: high) (also referred to as "first priority home judgment condition"), "energy usage per predetermined time" and "home judgment energy usage". (The amount of usage obtained by adding (+ α) the amount of error calculated statistically in advance to the minimum value within the period such as within the same month) "is used to make a home prediction.

In the home prediction using this home judgment condition (priority: high), if the "energy usage per predetermined time" is larger than the "home judgment energy usage", it is possible to stay at home at the predetermined time. On the other hand, if it is less than or equal to the "at-home judgment energy usage amount", it is judged that it is unlikely that the person is at home at the predetermined time (that is, there is a high possibility that he / she is absent). The probability of staying at home may be calculated based on the amount of deviation between the "energy usage per predetermined time" and the "home judgment energy usage". The larger the amount of divergence, the higher the probability of staying at home, and the smaller the amount of divergence, the lower the probability of staying at home.

In the above example, since the predetermined time is set to "15 minutes", "30 minutes", "1 hour", etc., it is possible to make a home prediction using the data of the energy consumption of high particle size. That is, it is possible to predict at home at a predetermined time such as "15 minutes", "30 minutes", and "1 hour".

The minimum value in the "home judgment energy usage amount" used in this home judgment condition (priority: high) indicates, for example, the standby power consumed by a constantly operating electric device such as a refrigerator. Further, the error amount indicates the additional power (power related to the allowable fluctuation) associated with the operation required for constant operation.

Next, as the next highest priority home judgment condition (priority: medium) (also referred to as "second priority home judgment condition"), home prediction is performed using the "past visit history and call history" for each user. There is something to do. This home-based determination condition (priority: medium) is a condition used for home-based prediction performed when the amount of energy used in the above-mentioned high-priority home-based determination condition cannot be obtained from each facility.

In the home prediction using this home judgment condition (priority: medium), a predetermined time such as several minutes before and after the visit time is referred to from the past visit history, the visit result at that time, and the like. Determine whether you are at home or not in the band (the same applies to the call history).

Then, as a home determination condition (priority: low) (also referred to as a "third priority home determination condition") having the lowest priority, there is one that performs home prediction using user attribute information. This home judgment condition (priority: low) is used when the energy consumption used in the home judgment condition (priority: high) shown above cannot be obtained from each facility, and in the home judgment condition (priority: medium). This is a home prediction performed when there is no past visit history or call history.

This home judgment condition (priority: low) is similar to that of a user who makes a home prediction (has a common point) without using the energy consumption of the user who makes a home prediction or the history of visits to facilities. It is a condition used to predict the user's home based on the attribute information.

In other words, when using the home judgment condition (priority: low), create (aggregate) and store "prediction reference data based on user attributes" that predicts home for each combination of user attribute information items. (See FIG. 6), home prediction is performed using this prediction reference data and at least a part of the user attribute information of the target user.

As described above, the predictive control process performed in the predictive control device 100 enables home prediction of all users by using home-based determination conditions according to the user, but by using a plurality of these home-based determination conditions. It is also possible to make a comprehensive home forecast. For example, it is possible to perform predictive control processing using a home determination condition (priority: high) and a home determination condition (priority: medium), and a home determination condition (priority: medium). It is also possible to perform predictive control processing using the home determination condition (priority: low).

In addition, in the result of judgment using a plurality of home judgment conditions, for example, the home prediction result judged by the home judgment condition (priority: high) and the home judgment condition (priority) to this home judgment condition (priority: high). If the home prediction result judged by taking into account (degree: medium) is different, review (update) the judgment criteria under the home judgment condition (priority: high), specifically, "home judgment energy consumption". ) Processing may be performed. As a result, even if the home prediction is performed by applying only the home determination condition (priority: high), it is possible to perform the home prediction with higher accuracy.

This will be described in detail. For example, as a result of home prediction using the home judgment condition (priority: high), the "past visit" of the home judgment condition (priority: medium) is in the time zone when it is judged to be "absent" (provisional judgment). If there is a result of being at home based on "history or call history", the visit result is the "home judgment energy usage" which is a parameter used for home prediction using the home judgment condition (priority: high). It is also possible to update by machine learning using the explained variable and the electric energy for each predetermined time (30 minutes, etc.) as explanatory variables. As a result, the home prediction determination logic using the home determination condition (priority: high) is updated (reviewed).

In the home determination in the prediction control device 100 shown above, an example using electric energy as the main energy is shown, but in addition to the main energy, the home determination process may be performed using the secondary energy. In this case, the predictive control device 100 can use the amount of secondary energy used as a variable for correcting the determination result of home determination using the amount of main energy (electrical energy) used, and also can use the main energy (main energy (electrical energy). It is also possible to determine at home by adding the amount of secondary energy used to the amount of electric energy used.

In this way, by using a plurality of different energies, it is possible to improve the accuracy of the determination result of the home determination by the prediction control device 100.

In addition to the above-mentioned home-based prediction function, the prediction control device 100 provides a visit route in which the order of visits is determined based on the information that can be visited to each facility obtained from the prediction result of the home-based prediction process. It also has a function to formulate a visit patrol plan by creating it (visit patrol plan formulation function). In this visit patrol plan, in addition to the prediction result of home prediction processing, map data (data for calculating distance and time), data on the number of people who can visit (number of people, skills, etc.), visit content (time required for the visit) Etc.) is used for planning.

Through this planning process, the visit patrol plan becomes a direct output (deliverables, outputs), while the home prediction result itself of each facility in the home prediction process becomes an indirect output, such as the home prediction result. It is possible to increase the confidentiality of highly confidential information.

By using this visit patrol plan to visit the facility, the visitor has a higher probability of visiting while the subject is at home, and the possibility of revisiting is reduced, resulting in an efficient visit. It will be possible.

Although the above description and FIG. 1 show an example in which the predictive control device 100 and the information management server 200 are configured by separate devices, these may be realized by a single device. Further, even if the configuration is such that the user management database 201 (customer DB), the energy data management database 202 (collected data DB), and the visit history management database 203 (visit history DB) are stored in the storage medium of the prediction control device 100. good. That is, the dotted line portion shown in FIG. 1 may be a single predictive control device 100. It is also possible to store the program of the predictive control processing performed by this single device in a storage medium (not shown) in the device and execute it in a calculation unit (CPU) or the like.

FIG. 2 is a diagram showing user information stored in a user management database included in the predictive control system according to the embodiment of the present invention.

The user information shown in FIGS. 2 (a) and 2 (b) consists of user identification information (user ID) for identifying the user and attribute information related to the user, and for convenience, FIGS. 2 (a) and 2 (b). It is described separately in (b) and may be composed of a single normalized information. In FIG. 2A, each attribute item of "name", "address", "family structure", and "contract energy plan" is stored in association with the user ID. The "contract energy plan" is a service plan provided for the energy contract used by the user. As an example shown in FIG. 2A, a user whose user ID is "user 001" has "○○ Taro" as attribute information for the "name" of the attribute item, and "address" of the attribute item. "Tokyo" is shown as the attribute information for "", and "single (male)" is shown as the attribute information for the attribute item "family structure", and the attribute item "contract energy plan". "Plan A" is shown as an attribute item for.

Subsequently, in FIG. 2B, each attribute item of "building use", "age of construction", and "owned equipment" is associated with the user ID. In addition, "owned equipment" indicates energy providing equipment according to the type of energy to be provided, and when the energy to be provided is "gas", "gas water heater" is indicated and provided. If the energy is "electric", then "electric water heater" is indicated.

In the example shown in FIG. 2B, the user whose user ID is "user 001" is shown as "detached house" as the attribute information for the attribute item "building use", and the attribute item "construction" is shown. "30 years or more" is shown as attribute information for "years", and "gas water heater" is shown as attribute information for "owned equipment" of the attribute item.

FIG. 3 is a diagram showing energy usage data stored in an energy data management database included in the predictive control system according to the embodiment of the present invention.

The energy usage data shown in FIG. 3 is associated with "energy usage per predetermined time" for each user. FIG. 3 shows an example of measuring the amount of energy used every "30 minutes" as a predetermined time for each date. In the example shown in FIG. 3, the energy measured at "11:00", "11:30", and "12:00" on "May 13, 2017" for the user whose user ID is "user 001". It indicates that the usage amount is "5", "5", and "4", respectively, and for the same user, "11:00", which is the same time of "May 14, 2017" on the next day, It is shown that the energy consumption measured at "11:30" and "12:00" is "6", "7", and "7", respectively.

That is, it is possible to grasp the energy usage status in the facility where the user is located with a high particle size from the energy usage data at short predetermined time intervals as shown in FIG.

Note that FIG. 3 shows the amount of energy used for any one of electricity, gas, oil (gasoline, kerosene), etc., but each energy (electricity, gas, oil (gasoline, kerosene)) ) May be managed individually.

FIG. 4 is a diagram showing visit history information stored in the visit history management database included in the predictive control system according to the embodiment of the present invention.

The visit history information in FIG. 4 is associated with each item of "visit date and time" and "visit result" with respect to the user ID of the user. In FIG. 4, for example, a visit to a facility having a user whose user ID is "User 002" at the visit date and time of "April 10, 2017 12:05" and the visit result of this visit date and time are shown. It shows that it was "absent". That is, it corresponds to the case where the presence or absence of the visit is stored in the visit history of the target person to the facility. It should be noted that the visit history to the facility where the user whose user ID is "user 003" is present does not store the presence or absence of the visit. That is, it corresponds to the case where the presence or absence of the visit is not stored in the visit history of the target person to the facility.

FIG. 5 is a schematic diagram illustrating the predictive control process performed by the predictive control device according to the embodiment of the present invention.

In FIG. 5, FIG. 5A shows a table obtained by cutting out a part of the “energy consumption per predetermined time” data shown in FIG. 3, and as an example, the user ID is “user 001”. It shows the energy consumption of two users (two persons) who are "users 002". In FIG. 5A, the energy consumption of the “user 001” measured at “11:00”, “11:30”, and “12:00” of “2017/05/13” in the instrument 400 is It is "5", "5", "4", and it is shown that the energy consumption of "user 002" is "4", "3", "3". In addition, the energy usage of "User 001" measured at "11:00", "11:30", and "12:00" at the same time of "2017/05/14" on the next day is "6", " 7 ”and“ 7 ”, indicating that the energy consumption of“ user 002 ”is“ 1 ”,“ 1 ”, and“ 5 ”. Of course, the table of FIG. 5A does not include users who cannot measure "energy consumption" with the instrument 400.

FIG. 5 (b) compares the "energy consumption per predetermined time" shown in FIG. 5 (a) with the "at-home judgment energy usage", and sets "1" when it is judged to be at home. It shows a table consisting of home prediction results (first home prediction results) in a state where "0 (zero)" is set when it is set and it is determined that the person is absent. In this example, the home prediction result when the "home judgment energy usage amount" is "2.5 (total value when the minimum value is" 2 "and the error amount is" 0.5 ")" is used. Shows.

FIG. 5 (c) takes into account the visit history to the home prediction result based on the comparison between the “energy usage per predetermined time” and the “home judgment energy usage” shown in FIG. 5 (b). It is a table which shows the home prediction result (second home prediction result) which was ensemble).

In FIG. 5 (c), the home prediction result at "2017/05/14 12:00" of "User 002" is "1 (at home)" with respect to the first home prediction result shown in FIG. 5 (b). Has been updated (changed) to "0 (absent)". This is because, for example, as the visit history information of "user 002" in FIG. 4, "absence" is shown at "12:05" of "2017/4/10", so that "predetermined time" in this time zone It was updated (changed) by considering (ensemble) the "energy consumption per unit" and the past visit history, and judging that there is a high possibility of "absence". Of course, the "energy consumption per predetermined time" and the past visit history may be determined by applying a predetermined weight to each.

These FIGS. 5 (b) and 5 (c) show the results of home prediction based on whether or not the person is at home (“100%” or “0%”) for each predetermined date and time and for each time zone in the prediction control process. be. Therefore, the home prediction result as shown in FIGS. 5 (b) and 5 (c) may be output as a “home status list”.

Subsequently, FIG. 5D shows the home prediction result based on the home probability.

In FIG. 5 (d), FIG. 5 (d-1) shows the probability of staying at home per predetermined time (per 30 minutes), and FIG. 5 (d-2) shows three time zones (1). It shows the probability of staying at home in each time zone divided into night, morning, and afternoon).

5 (d-1) and 5 (d-2) are one of the “first home prediction result” of FIG. 5 (b) or the “second home prediction result” of FIG. 5 (c). By statistically processing both and the visit history information of FIG. 4, the home prediction result based on the home probability is shown. For example, it can be calculated by applying a predetermined value based on the home prediction result to the number of times of staying at home or the number of times of absence with respect to the total number of visits.

FIG. 6 is a diagram showing predictive reference data used in predictive control processing performed by the predictive control device according to the embodiment of the present invention.

FIG. 6 shows prediction reference data showing the probability of staying at home in each time zone (in units of 30 minutes) for a combination of user attribute information items. As the item of the user attribute information of the prediction reference data shown in FIG. 6, "building use" and "age of building" are shown, and the home prediction of the user's facility having the value (home-based probability) in this item as the user attribute information. Will be used for.

At this time, what kind of combination should be taken, that is, what kind of attribute information should be classified and patterned, can be defined according to the classification form required for business, and each attribute information. It is possible to search and define by a statistical method such as the correlation and correlation distribution of.

Since it is possible to predict the user's home at each time zone based on the user attribute information in FIG. 6, the energy consumption of the facility where the user is located (data used for the first priority home determination condition) and Even if the visit history (data used for the second priority home determination condition) and the like are not registered in the predetermined database in advance, the user attribute information is registered in the user management database 201 based on this prediction reference data. Home prediction is possible for all users.

FIG. 7 is a flowchart showing a flow of predictive control processing performed by the predictive control device according to the embodiment of the present invention.

In FIG. 7, the prediction control device 100 makes a home prediction request from a “home prediction request source” such as an input device (keyboard, mouse, etc.) connected to a predetermined terminal (PC, mobile) or the self-prediction control device 100. Processing is started by receiving. This home prediction request includes information for designating a user who performs home prediction as a "home prediction condition" for performing home prediction. For example, it is specified to perform home prediction for all registered users, or to perform home prediction for one or more designated users.

Upon receiving this home prediction request, the prediction control device 100 extracts user information about the user included in the "home prediction condition" (S701). The user information at this time includes a user ID that identifies the user and user attribute information.

Subsequently, the predictive control device 100 determines whether or not the data of the "first priority home determination condition" for the "1 user" in this user information is stored (S702). As shown above, this "first priority home determination condition" is a condition for performing home prediction based on the "energy consumption per predetermined time", and therefore, the "first priority home determination condition". Data indicates the amount of energy used received from the facility where "1 user" is located. Therefore, the process of step 702 is a process of determining whether or not the energy consumption amount is stored for "1 user".

When it is determined by this determination process that the data of the "first priority home determination condition" is stored by storing the energy consumption (YES in S702), the prediction control device 100 uses this "first". Based on the "priority home-based determination condition", the home / absence determination process is performed in each time zone (S703). In this determination process, it is determined whether the user is "at home" or "absent" by using the "energy usage per predetermined time" of "1 user" and the "home determination energy usage". If it is not determined that the data of the "first priority home determination condition" is stored (NO in S702), the process proceeds to step 705 described later.

Then, the prediction control device 100 creates a first home prediction result by the home / absence determination process in each time zone based on this "first priority home determination condition" (S704). As shown in FIG. 5B, the first home prediction result shows, for example, the result based on the magnitude relationship between the "energy consumption per predetermined time" and the "home judgment energy usage". ..

Subsequently, the prediction control device 100 determines whether or not the data of the "second priority home determination condition" for the "1 user" is stored (S705). As shown above, this "second priority home determination condition" is a condition for making a home prediction based on the past visit history and call history for "1 user", and therefore the "second priority home determination condition". The data of "priority home determination condition" indicates a visit history and a call history. Therefore, the process of step 705 is a process of determining whether or not the visit history and the call history are stored for "1 user".

When it is determined by this determination process that the data of the "second priority home determination condition" is stored by storing the data of the past visit history and the call history (YES in S705), the prediction control device. 100 performs home / absence determination processing in each time zone based on this “second priority home determination condition” (S706). In this determination process, it is determined whether the user was "at home" or "absent" for each time zone using the past visit history and call history of "1 user".

Subsequently, the prediction control device 100 creates a second home prediction result by the home / absence determination process in each time zone based on this "second priority home determination condition" (S707). The second home prediction result shows information on whether the user was "at home" or "absent" such as past visit history and call history, and the first home prediction result is created by the process of step 704. If so, the first home prediction result is shown by adding (ensemble) information on whether it was "at home" or "absent" such as past visit history and call history.

Then, the prediction control device 100 calculates the user's home probability based on at least one of the first home prediction result and the second home prediction result (S708). In the process of step 702, when it is determined that the data of the "first priority home determination condition" is stored, the "first home prediction result" created in the process of step 704 and the "first home prediction result" created in the process of step 707. The user's home probability is calculated using the "second home prediction result". Further, in the process of step 702, if it is not determined that the data of the "first priority home determination condition" is stored, the "first home prediction result" is not created, so that the data was created in the process of step 707. The user's home probability is calculated using the "second home prediction result".

After the user's home probability is calculated by the process of step 708 in this way, the predictive control device 100 subsequently stores the data of the "first priority home determination condition" in the process of step 702. If it is not determined (NO in S702), the user attribute information of "1 user", which is the data of the "third priority home determination condition", is read out (S710).

Since this "third priority home determination condition" is a condition for performing home prediction using the user attribute information, the data of the "third priority home determination condition" indicates the user attribute information.

Then, the prediction control device 100 identifies the user's home probability from the prediction reference data as shown in FIG. 6 based on the user attribute information (S711). As shown in FIG. 6, this specifying process is a process of specifying one of the home probabilities as the home probability of the "1 user" because the home probability is shown as the prediction reference data.

Then, the prediction control device 100 creates a third home prediction result indicating the home probability in each time zone based on the user attribute information of "1 user" and the data of the "third priority home determination condition" ( S712). This creation process is a process that is always performed when the first home prediction result is not created, regardless of whether or not the second home prediction result is created by the process of step 707. That is, if the first home prediction result is not created based on the home determination condition having the highest priority (first priority home determination condition), the prediction control device 100 performs home prediction based on the user's attribute information. It shows that.

Further, when it is determined by the process of step 708 that the user's home probability is calculated and then the data of the "first priority home determination condition" is stored by the process of step 702 (YES in S702), that is, , The "first home prediction result" is created based on the data of the "first priority home judgment condition", and the "second home prediction result" is created based on the data of the "second priority home judgment condition". If created, the predictive control device 100 performs the process in step 713 described later.

On the other hand, when it is not determined by the process in step 705 that the data of the "second priority home determination condition" is stored (NO in S705), that is, whether or not the user has visited the facility. Is not stored, and when it is determined that the "first home prediction result" has not been created by the process in step 702 (NO in S702), the prediction control device 100 is described above. Each process after the process of step 710 is performed. From this, even if neither the data of the "first priority home determination condition" nor the data of the "second priority home determination condition" is stored for the "1 user", this "1 user" It is possible to make a home prediction based on the attribute information of "1 user".

Further, in the process of step 705, when it is not determined that the data of the "second priority home determination condition" is stored (NO in S705), that is, the presence or absence of a visit is stored in the visit history of the user to the facility. If it is not done and it is determined in the process of step 702 that the data of the "first priority home determination condition" is stored (YES in S702), that is, there is one user. When the energy usage data received from the facility is stored, the prediction control device 100 calculates the user's home probability based on the first home prediction result (S708). Then, the predictive control device 100 performs the process in step 713 described later.

The prediction control device 100 determines whether or not the home prediction has been completed for all users by performing the processes of steps 702 to 712 as described above for all users (S713).

When the home prediction for all the users specified in the home prediction request is completed by performing the processing as shown in steps 702 to 712 (YES in S713), the prediction control device 100 subsequently determines each user. The home-based probability is aggregated, and a "home-based status list" showing the home-based prediction result associated with the home-based probability for each time zone is created for each user (S714).

Then, the prediction control device 100 generates "prediction reference data" based on the user attribute information of each user (S715). The prediction reference data generated in step 715 is stored in the storage unit of the prediction control device 100 and can be registered in any database. Further, it is possible to output the "home status list" to the terminal connected to the predictive control device 100 and output it to the display device of this terminal or a predetermined medium.

This prediction reference data is stored when performing home prediction for a user who does not store either the data of the "first priority home determination condition" or the data of the "second priority home determination condition" in step 710. It is the data read from the unit and used for home prediction.

From the above, the predictive control process in the predictive control device 100 of the present invention predicts the home situation with high accuracy based on the predictive judgment conditions according to each user for the designated user or all users (home prediction). ) Is possible.

That is, even if the energy consumption data from the facility where the user is located cannot be obtained, or even if the visit to the facility where the user is located is the first time, the predictive control device 100 of the present invention provides high accuracy. It is possible to predict the home situation of.

The embodiment shown above is an embodiment of the present invention, and is not limited to these examples, and can be appropriately modified and implemented without changing the gist thereof.

100 Predictive control device 200 Information management server 201 User management database 202 Energy data management database 203 Visit history management database 300 Information communication control device 400 Instrument 500 Home system 501 Home energy management device 502 Energy utilization device

Claims (7)

Attribute information storage means for storing attribute information about the target person to be visited by the visitor,
The energy information consumed in the facility where the subject can stay at home is measured by the instrument, and the acquisition means for acquiring the energy information from the instrument and the acquisition means.
When the energy information of the facility can be acquired, the home status of the target person at the time when the energy information is measured is predicted based on the energy information acquired by the acquisition means and the information for determining the energy information. 1 home situation prediction means and
A prediction reference used for home prediction of another target person to which the attribute information is associated based on the home status predicted by the first home status prediction means and the attribute information related to the target person in the home status. How to create data and
A second method of predicting the home status of the target person by referring to the prediction reference data based on the attribute information of the target person stored in the attribute information storage means when the energy information of the facility cannot be acquired. A predictive control device characterized by being equipped with a home situation predicting means. Further equipped with a visit history storage means for storing the visit history of the target person to the facility,
The visit history includes a case where the presence or absence of a visit to the facility of the subject is memorized and a case where the presence or absence of the visit is not memorized.
The predictive control device according to claim 1. The first home-based situation prediction means is
2. The predictive control device described. The second home situation prediction means is
When the energy information in the facility cannot be acquired and the presence or absence of a visit is not stored in the visit history of the subject to the facility by the visit history storage means, the home status of the subject is predicted. The predictive control device according to claim 2, wherein the predictive control device is characterized in particular. Based on the attribute information of the target person who predicted the home situation by at least one of the first home situation prediction means and the second home situation prediction means, the home status of another target person having the attribute information can be determined. Further equipped with a generation means for generating predictable predictive reference data,
The second home situation prediction means is
Claim 1 is characterized in that the home status of the target person is predicted by using the prediction reference data generated by the generation means based on the attribute information of the target person for the facility visited by the visitor. Alternatively, the predictive control device according to claim 4. Attribute information storage means for storing attribute information about the target person to be visited by the visitor,
A prediction control device including a home prediction means for predicting the home status of the target person, and
Energy utilization equipment installed in the facility visited by the visitor and operating using energy,
It is equipped with a measuring means for measuring energy information consumed by the energy-using device.
The home prediction means is
When the energy information of the facility can be acquired , the home state of the target person at the time when the energy information is measured is predicted based on the energy information measured by the measuring means and the information for determining the energy information. Further equipped with a first home situation prediction means,
The predictive control device is
A prediction reference used for home prediction of another target person to which the attribute information is associated based on the home status predicted by the first home status prediction means and the attribute information related to the target person in the home status. How to create data and
When the energy information cannot be measured at the facility, the home status of the target person is predicted by referring to the prediction reference data based on the attribute information of the target person stored by the attribute information storage means. A prediction control system characterized by further being equipped with a home prediction means. Attribute information about the target person to be visited by the visitor is stored in the storage medium,
The step in which the predictive control device measures the energy information consumed in the facility where the subject can stay at home.
When the energy information of the facility can be acquired , the predictive control device predicts the home state of the target person at the time when the energy information is measured, based on the measured energy information and the information for determining the energy information. Step,
A step in which the prediction control device creates prediction reference data to be used for home prediction of another target person to which the attribute information is associated based on the predicted home status and the attribute information related to the target person in the home status. ,
A step in which the prediction control device predicts the home status of the target person by referring to the prediction reference data based on the attribute information of the target person when the energy information cannot be measured in the facility.
Predictive control method consisting of.

Images