JP2015153156A - Device, program and method capable of allowing for consistent estimation of living situation based on power consumption - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a living situation estimation device allowing for consistent estimation of a plurality of living situations based on power consumption.SOLUTION: A living situation estimation device includes: a plurality of living situation estimation means that estimate a living situation for each of a plurality of items of living situation to be estimated based on data of user's power consumption; estimation result check means that determines consistency between estimation results of living situation for each of the plurality of items based on a living situation rule prescribing consistency between living situations; evaluation value management means that maintains evaluation values calculated based on degrees of estimation accuracy in each living situation estimation means; and estimation result correction means that, for a pair of estimation results determined to be inconsistent, adjusts one or more estimation results of the estimation results of this pair based on an evaluation value of estimation results of this pair to be consistent, calculates total evaluation values, and compares the calculated total estimation values to determine a pair of corrected estimation results.

Description

  The present invention relates to a user information estimation technique that analyzes an amount of power consumed by a user and estimates predetermined information about the user.

  2. Description of the Related Art In recent years, attention has been focused on systems that provide support and advice related to life, such as household management support and power saving advice, in accordance with a user's life situation, for example, information on home stay and bedtime. Here, as one method for grasping the user's life situation, it is possible to use the amount of power consumed by the user. The advantage of using this power consumption is that the living situation can be estimated by a non-intrusive method without entering the living space such as the user's house.

  As a technique for estimating a living situation by using power consumption, for example, Patent Document 1 acquires power consumption using a power meter installed near a service line of a power consumer, and the intention of a resident A technique for estimating a living situation such as safety of a resident, presence / absence, sleeping / non-sleeping by determining whether or not a typical electric device is used is disclosed. Here, in order to improve the estimation accuracy of the living situation, the relationship between the resident's past living situation and the power consumption is utilized.

  Further, Patent Document 2 discloses an action determination system that can reliably determine whether the person being watched over is at home or absent. Specifically, using the operational status of each electrical device and the functional relationship for estimating the home state according to the attributes such as the age of the person being watched over, the current home state of the person being watched over is estimated. Yes.

JP 2011-43984 A JP 2010-182072 A

  The conventional techniques as described in Patent Documents 1 and 2 only estimate a specific living situation. For example, the home situation is estimated from the power consumption, and only a contrivance is made to improve the specific estimation accuracy.

  However, in reality, the living conditions to be acquired for use in services to users often include many items. For example, it is possible to form and provide more appropriate services after comprehensively judging the living situation over many items such as whether you are sleeping, whether you are cooking, whether you are taking a bath, etc. It is required to do. Here, although it is possible to estimate individual living situations using conventional technology, there is no technology for comprehensive judgment of estimation results, and for example, inconsistent estimation results such as sleeping and absent are output. There is a possibility of being.

  Then, an object of this invention is to provide the living condition estimation apparatus, program, and method which can perform the consistent estimation about several living conditions based on power consumption.

According to the present invention, an apparatus for estimating a user's life situation based on a user's power consumption,
A plurality of living situation estimation means for estimating a living situation for each of a plurality of living situation items to be estimated based on data of a user's power consumption;
An estimation result confirming means for judging the consistency between the estimation results of the living situation in each of the plurality of living situation items based on the living situation rule that determines the consistency between the plurality of living situations;
An evaluation value management means for holding an evaluation value calculated based on the estimation accuracy in each of a plurality of living situation estimation means;
For a set of estimation results determined to be inconsistent, based on the evaluation value related to the estimation result of this set, one or more of the estimation results of this set are changed to be in a consistent state And calculating the total evaluation value, and comparing the calculated total evaluation values with each other, an estimation result correcting unit that sets the set of the estimated results after the change with the highest evaluation as a set of the corrected estimation results. A user life situation estimation device is provided.

In this user life situation estimation device, the evaluation value management means is:
(A) the above estimation accuracy;
(B) It is also preferable to hold the evaluation value calculated based on the reliability score of the living situation estimation result by the living situation estimating means.

Further, as a specific embodiment of the user life situation estimation apparatus of the present invention, the life situation estimation means sets a condition related to the power consumption for the life situation item to be estimated, and the power consumption satisfies this condition. Based on whether or not
The evaluation value management means calculates the reliability score using the power consumption amount in the estimation target time zone or time and the value that defines the condition used for the estimation, and for each estimation result of the living situation, Calculate the evaluation value based on the product of (estimation accuracy of estimation means) and (calculated reliability score)
The estimation result correcting means, for a set of estimation results determined to be inconsistent, based on a calculated evaluation value relating to the estimation result of this set, one or a plurality of estimation results of the estimation results of this set The total value of the evaluation values related to the unestimated estimation results in the case of changing to a consistent state is calculated as the total evaluation value, and the set of estimation results after the change that maximizes the total evaluation value is corrected. It is also preferable to use a set of estimation results.

  Furthermore, as another embodiment of the user life situation estimation apparatus of the present invention, at least one of a plurality of life situation estimation means estimates a life situation for a life situation item related to the number of people staying in a user's household area. It is also preferable that at least one of the remaining living state estimating means estimates the living state for the living state item related to the living behavior of the user in the household area.

  As still another embodiment of the user life situation estimation apparatus according to the present invention, the estimation result confirmation means selects a life situation rule to be used when judging consistency based on user information related to the user's life. It is also preferred to do or change.

  Furthermore, as still another embodiment of the user life situation estimation apparatus of the present invention, at least one of the plurality of life situation estimation means is not only for the user's total power consumption data or for the total power consumption. It is also preferable to estimate the living situation based on power consumption data by a specific device used by the user instead of the data.

According to the present invention, there is also provided a life situation estimation program for causing a computer mounted on a device that estimates a user's life situation based on a user's power consumption to function.
A plurality of living situation estimation means for estimating a living situation for each of a plurality of living situation items to be estimated based on data of a user's power consumption;
An estimation result confirming means for judging the consistency between the estimation results of the living situation in each of the plurality of living situation items based on the living situation rule that determines the consistency between the plurality of living situations;
An evaluation value management means for holding an evaluation value calculated based on the estimation accuracy in each of a plurality of living situation estimation means;
For a set of estimation results determined to be inconsistent, based on the evaluation value related to the estimation result of this set, one or more of the estimation results of this set are changed to be in a consistent state The computer is used as an estimation result correction unit that calculates the total evaluation value in the case of the above, and compares the calculated total evaluation values with each other to make the set of the estimated results after the change having the highest evaluation the set of the corrected estimation results. A living situation estimation program is provided to make the functioning of

According to the present invention, there is further provided a living situation estimation method for estimating the living situation of the user based on the user's power consumption amount,
A first step of estimating a living situation for each of a plurality of living situation items to be estimated based on data of a user's power consumption;
A second step of determining the consistency between the estimation results of the living situation in each of the plurality of living situation items, based on the living situation rule that determines the consistency between the plurality of living situations;
A third step of holding an evaluation value calculated based on the estimation accuracy in each estimation in the first step;
For a set of estimation results determined to be inconsistent, based on the evaluation value related to the estimation result of this set, one or more of the estimation results of this set are changed to be in a consistent state And calculating the total evaluation value in the case of the above, and comparing the calculated total evaluation values with each other, a fourth step of setting the set of the estimated result after the change with the highest evaluation as the set of the corrected estimation result, A living situation estimation method is provided.

  According to the living situation estimation apparatus, program, and method of the present invention, it is possible to acquire consistent estimation results for a plurality of living situations based on the power consumption.

It is a schematic diagram in one Embodiment of the living condition estimation system which concerns on this invention. It is a functional block diagram which shows the function structure in one Embodiment of the living condition estimation apparatus by this invention. It is a graph which shows one Example of the power consumption of a user household. It is a table which shows the estimation conditions of the living condition in a present Example. It is a graph which shows the living condition rule in a present Example. It is a graph which shows the estimation result of the living condition in a present Example. It is a program which shows one Example which calculates a total evaluation value and correct | amends the group of an estimation result. It is a table which shows the evaluation value about each estimation result in a present Example.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a schematic diagram of an embodiment of a living situation estimation system according to the present invention.

  As shown in FIG. 1, the living situation estimation system 1 according to the present embodiment is installed in a provider communication network 5 and estimates a user's “living situation” based on the user's power consumption. including. Here, the user is one household (user household) or a member of the household in the present embodiment.

  In addition, “life situation” refers to the state or behavior of a user in a household area (home) that changes depending on the living behavior of the resident user, such as at home, sleeping, cooking, or bathing. Point to. Therefore, for example, an in-home environment situation where the temperature inside the house also rises due to an increase in the temperature outside the house does not apply to this living situation.

  Similarly, according to FIG. 1, the power consumption amount of the user (one household) is the living situation from the home gateway (HGW) 4 installed at the home which is the user's household area in this embodiment through the operator communication network 5. It is transmitted to the estimation device 1. Here, the power consumption may be measured by the smart meter 3, and the measured value may be output to the HGW 4 via, for example, a HEMS (Home Energy Management System). Moreover, as a change mode, the smart meter 3 can transmit power consumption to MDMS (Meter data Management System) (not shown), and the living situation estimation apparatus 1 can acquire the power consumption from this MDMS. It is.

  The carrier communication network 5 can be a fixed access network such as an optical fiber network or an ADSL (Asymmetric Digital Subscriber Line). In addition, wireless LAN (Local Area Network) such as Wi-Fi (registered trademark), WiMAX (Worldwide Interoperability for Microwave Access), LTE (Long Term Evolution) or 3G (3rd Generation) or other wireless access network may be used. Is possible.

  The smart meter 3 measures the amount of power consumed by the user at home, and outputs a measured value of the amount of power consumed per unit time to the HGW 4 via, for example, HEMS. Here, instead of the smart meter, a CT sensor installed on a distribution board, a clamp meter (overhead ammeter), or an outlet tap (tap wattmeter) installed at an outlet may be used. In this case, the power consumption amount of each distribution in the house or the individual power consumption amount of the home appliance is measured by a CT sensor, an outlet tap, etc., and the measured value of the power consumption amount is output to the HGW 4.

  The HGW 4 is a device that relays and controls communication between a home network provided in a home, a set top box (not shown) (STB), and the like, and the carrier communication network 5. For example, a personal computer (PC) is connected to the home network via a wireless access point (AP). The HGW 4 further transmits the measurement value of the power consumption per unit time acquired from the smart meter 3 to the living situation estimation device 1 via the business operator communication network 5.

  The user information management device 2 is installed, for example, in the carrier communication network 5 and receives user information declarations from each user (household) in the form of, for example, an answer to a questionnaire survey. Store and manage in association with. For example, the user inputs user information such as the number of members of his / her household, sex, age and annual income, and household address to the PC, and transmits the information from the HGW 4 to the user information management apparatus 2 via the carrier communication network 5. May be. The living situation estimation apparatus 1 acquires user information associated with the user ID from the user information management apparatus 2.

Specifically, the living situation estimation device 1
(A) Based on the user's power consumption data, “multiple living status items”, for example, the number of people at home (the number of people staying in the household area) is 0, 1 or 2 Or whether you are sleeping, whether you are taking a bath, and whether you are cooking,
(B) Based on the “living situation rule” that determines consistency between a plurality of living situations, the consistency between “estimated results of living situations” in each of the above “multiple living situation items” is determined. ,
(C) holding an “evaluation value” calculated based on the “estimation accuracy” in each of the means for performing a plurality of estimations in (a),
(D) For a group of “estimation result of living situation” determined to be inconsistent, one of the “estimation result of living situation” of this group based on the “evaluation value” related to the estimation result of this group Calculate the “total evaluation value” when changing one or more estimation results to a consistent state, and compare the calculated “total evaluation values” with each other to obtain the estimated result after the change with the highest evaluation. Is a corrected “estimation result of living situation”.

  As described above, in the living situation estimation apparatus 1, when it is judged that there is no consistency after judging the consistency of the "living situation estimation result", a part of this group is changed and corrected. “A set of estimation results” can be provided. As a result, it is possible to estimate a plurality of consistent living situations based on the power consumption.

  Further, since the living situation estimation apparatus 1 independently derives and handles the “living situation estimation result”, when adding a “living situation item” to be estimated, the living situation estimation means corresponding to the item is added to the apparatus. In addition, it is possible to respond to the addition by simply adding a rule to the “living situation rule”. That is, it is not necessary to change the existing living situation estimation unit. As a result, it is possible to easily cope with a case where it is desired to estimate a comprehensive life situation across multiple items.

  Here, it is also preferable that the “evaluation value” described in the above (c) is calculated based on the “estimation accuracy” and the “reliability score” of the “living situation estimation result”. Thereby, it is possible to more appropriately correct the “estimation result of the living situation”. Such a method of calculating the “evaluation value” will be described later as one embodiment with reference to FIG.

  FIG. 2 is a functional block diagram showing a functional configuration in an embodiment of the living situation estimation apparatus according to the present invention.

  According to FIG. 2, the living situation estimation apparatus 1 includes a communication interface unit 101, a power consumption data storage unit 102, and a processor memory. Here, the processor memory realizes its function by executing a program that causes the computer mounted in the living situation estimation apparatus 1 to function. In addition, the processor memory includes N (N is a natural number) living state estimating units, a first living state estimating unit 111a, a second living state estimating unit 111b,..., And an Nth living state estimating unit. 111n, the estimation result confirmation unit 112, the evaluation value management unit 113, and the estimation result correction unit 114 function. In addition, according to FIG. 2, the flow of the process which connected each function structure part with the arrow is understood also as one Embodiment of the living condition estimation method by this invention.

The communication interface unit 101
(A) The measured value data of the power consumption is received from the smart meter 3 or the HGW 4 installed in the user's home (household area), and the user ID and the collected user information are received from the user information management device 2. Receive pairs, and
(B) Transmit (corrected) “estimation result of living situation” for the estimation target user output from the estimation result correction unit 114 to an external notification destination. For example, the estimation result can be presented via a website or application.

  The power consumption data accumulation unit 102 accumulates and manages power consumption data generated from the measurement value of the power consumption received via the communication interface unit 101. For example, a 24-dimensional or 48-dimensional power consumption vector is generated from the power consumption in the hour or 30 minutes constituting a day, and M corresponding to each day in M days (M is a natural number) May be obtained, stored, and managed. Note that the power consumption data may be accumulated outside the apparatus 1, for example, in a storage in the user's home.

  The first to Nth living situation estimation units 111a to 111n estimate the living situation for each of a plurality of living situation items to be estimated based on the user power consumption data, and determine the consistency of the estimation result. To the unit 112c. That is, each estimation part 111a-111n is an estimator about one mutually different living situation item. Specific estimation methods themselves in each of the estimation units 111a to 111n include, for example, estimation of presence / absence by simple threshold determination for power consumption, and power consumption data divided every hour or every 30 minutes. Various known methods such as a vectorization method, a clustering method, a random forest method, and a logistic regression analysis method can be applied.

  As described above, each of the first to Nth living situation estimation units 111a to 111n may have a configuration independent of each other. Thereby, when it becomes necessary to add a living situation item to be estimated, a new living situation estimation unit can be added to easily cope with it. Actually, the living situation items to be estimated may have various patterns depending on the use destination of the estimation result, but the estimation unit of the present embodiment is configured to facilitate the addition of such living situation items. .

  In addition, at least one of the first to N-th living state estimating units 111a to 111n (for example, the third living state estimating unit 111c in FIG. 2) is not only data of the total power consumption of the user or all of them. Instead of the power consumption data, the living situation may be estimated based on the power consumption data by a specific device used by the user. For example, when power consumption data is acquired for a power distribution system in which home appliances such as air conditioners and IH (Induction Heating) cookers are independently connected to each other, based on this data, It is also possible to estimate whether cooking is in progress.

  Here, an example of living state estimation will be shown. In this embodiment, it is assumed that user information indicating that the number of household members is two in one user household is acquired. In addition, as the first to sixth living state estimation units 111a to 111f, “0 person staying at home (whether or not)”, “1 person staying at home (whether or not)”, “two people at home (whether or not) ) ”,“ Sleeping (whether) ”,“ taking a bath (whether) ”and“ cooking (whether) ”are used to estimate each of the six living situation items.

  As described above, at least one of the living situation estimation units estimates a living situation with respect to a living situation item related to the number of people living at home (the number of people staying in the household area). It is also preferable that at least one of these estimates a living situation for a living situation item related to a living behavior of a user in a household area. Accordingly, it is possible to perform estimation with higher reliability more reliably by determining the consistency between these two types of living situations that are likely to be inconsistent with each other. In addition, since the number of household members is two, even if an estimation unit that performs home determination of three or more people is installed, it is preferable that an estimation unit that can obviously make a contradictory determination not function.

  FIG. 3 is a graph showing an example of the power consumption of the user household. FIG. 4 is a table showing living condition estimation conditions in the present embodiment.

  FIG. 3 shows the power consumption amount for 30 minutes for the main power in the time zone every 30 minutes in a certain day in the user household of the present embodiment. It is understood that the power consumption is prominently increased in the time zone including 7:30 (7: 15-7: 45) and the time zone including 19:00 (18: 45-19: 15) .

  Moreover, as shown in FIG. 4, the estimation method in each of the first to sixth living situation estimation units 111a to 111f in the present embodiment is such that the power consumption amount and the time (time zone) have predetermined conditions. This is a method of determining whether or not the living situation item is true, that is, estimating the living situation, based on whether or not it is satisfied (specifically, whether or not it is within a predetermined threshold range). For example, if the power consumption is 5 kWh at 21:00 (including time zone (20:45 to 21:15)), it is determined as “sleeping”.

  In addition, the estimation based on the data of the power consumption amount in the living situation estimation unit includes the estimation based on the processed power consumption amount, for example, the estimation based on the difference from the previous power consumption amount, and further the power consumption amount It is also preferable to include estimation using a related amount such as measurement time (time zone) and standby power amount.

  Returning to FIG. 2, the estimation result confirmation unit 112 determines the consistency between the estimation results of the living situation in each of the plurality of living situation items based on the living situation rule. In the present embodiment, the estimation result confirmation unit 112 includes a living situation rule management unit 112a, a user information management unit 112b, and a consistency determination unit 112c.

  The living situation rule management unit 112a holds and manages a living situation rule in which consistency between a plurality of living situations is negotiated. The living situation rule itself may be generated manually, for example, according to the assumed living situation item.

  FIG. 5 shows a living situation rule in the present embodiment. According to FIG. 5, ten rules R1 to R10 are listed as living situation rules for determining consistency. In each rule, a value 1 (true) or a value 0 (false) is set for each living situation item. For example, in the rule R3, “0 at home” and “sleeping” are 1, and 0 in other life situation items. This means that when the true determination is made for both the living status items “at home for 0” and “sleeping”, these determinations are not consistent. That is, the rules R1 to R10 define combinations of living situation items that are not consistent with each other.

  Returning to FIG. 2, the user information management unit 112 b acquires, accumulates and manages a set of the user ID and user information received via the communication interface unit 101. Here, the user ID is an identification number of a user household or household member. Further, in the present embodiment, the user information is static information that does not change with time, such as the living situation, such as the number of members of the household, sex, age and annual income, and household address.

  Here, the living situation rule management unit 112a may select or change the “living situation rule” used for the consistency determination based on the user information related to the user's life notified from the user information management unit 112b. preferable. For example, when user information indicating that there is no bathroom at home is obtained, the estimation unit for estimating whether or not “bathing” is not installed or used, and “bathing” Can be excluded.

  The consistency determining unit 112c inputs the estimation result (estimation result set) of the living situation in each estimating unit from the first to Nth living situation estimating units 111a to 111n. Next, based on the “life situation rule”, the consistency between the life situation estimation results in each of the plurality of life situation items is determined. Hereinafter, an embodiment for determining the consistency will be described based on the embodiment shown in FIGS.

  FIG. 6 is a graph showing the estimation result of the living situation in the above-described embodiment. The estimation results in the figure are obtained based on the power consumption shown in FIG.

  FIG. 6A shows the result of determining the authenticity in each time zone for each of “at home for 0”, “at home for 1” and “at home for 2”. In the graph of FIG. 8, “2 people staying at home” does not appear because it was a false determination in all time zones. FIG. 6B shows the result of determining the authenticity in each time zone for each of “sleeping”, “taking a bath”, and “cooking”.

  As shown in FIGS. 6 (A) and 6 (B), since there are a plurality of life status items to be estimated in the first place, the estimation results overlap depending on the time zone. For example, at time 7:30 and 19:00, “1 person A set of estimation results of “at home”, “bathing” and “cooking” is obtained. Here, when the consistency in the set of these estimation results is examined using the “life situation rule” shown in FIG. 5, the rule R9 is just applied to these estimation results. That is, it is pointed out that rule R9 is not consistent with this set of estimation results.

  Returning to FIG. 2, the consistency determination unit 112 c performs the consistency determination as described above, and outputs the determination result to the estimation result correction unit 114.

The evaluation value management unit 113
(A) "Estimation accuracy" indicating the estimation performance in each of the first to Nth living situation estimation units 111a to 111n;
(B) The “evaluation value” calculated based on the “reliability score” of the estimation result of the living situation by each of the first to Nth living situation estimation units 111a to 111n is held. Here, the evaluation value management unit 113 itself may calculate such an “evaluation value”.

Here, “estimation accuracy” will be described in detail in a later embodiment. For example, for a living situation estimation unit that estimates “sleeping (or not)”, the accuracy of the estimation result by this estimation unit is It is data obtained from past results and results such as 80%. Also, the “reliability score” will be specifically described in the following example, but when the living situation is estimated under the conditions shown in FIG.
(A) power consumption in the estimation target time zone or time;
(B) It is calculated using a value that defines a condition used for estimation (for example, a power consumption value that defines a condition range used for estimation). The evaluation value management unit 113 may calculate an “evaluation value” for each life situation estimation result based on the product of the “estimation accuracy” of the living situation estimation unit and the calculated “reliability score”. preferable.

The estimation result correction unit 114 inputs an “evaluation value” related to the estimation result of this set from the evaluation value management unit 113 for the set of estimation results determined to be inconsistent,
(A) Based on the input “evaluation value”, calculate the “total evaluation value” when one or more of the estimation results in this set are changed to a consistent state,
(B) By comparing the calculated “total evaluation values”, a set of estimation results having the highest evaluation (maximum total evaluation value) is determined as a set of corrected estimation results.

Here, the correction means that, as shown in (a) above, a part of each estimation result is changed so that there is consistency as a whole estimation result. The estimation result correction unit 114 determines a set of estimation results determined to be inconsistent based on the calculated “evaluation value” related to the estimation result of the set,
(A) Calculate the total value of the “evaluation values” related to the unchanged estimation results when one or more of the estimation results of this set is changed to a consistent state. Total evaluation value "
(B) It is also preferable to set a set of corrected estimation results that maximizes the calculated “total evaluation value” as a corrected set of estimation results.

  Hereinafter, an embodiment in which a “total evaluation value” is calculated and a set of estimation results is corrected will be described based on the embodiments shown in FIGS. 3 to 6.

  FIG. 7 is a source code of a program showing an embodiment for calculating a total evaluation value and correcting a set of estimation results.

  In the source code of the present embodiment shown in FIG. 7, an evaluation value is calculated for each estimation result, and the estimation result is inverted for a combination that increases the total evaluation value. The number of elements to be reversed is increased, and the correction process is completed when the estimation result as a whole does not violate the living situation rules. At this time, a total evaluation value is calculated for a combination of living situations that satisfies the rule condition Rj, and the set of estimation results is corrected to a combination of combinations that maximizes the total evaluation value.

Here, the functions shown in the source code of FIG. 7 will be described below.
(A) get_common (argument C: original estimation result, argument Rj: living situation rule)
Processing content: The element C of C described in Rj is extracted for the original estimation result C and the living situation rule Rj determined to be inconsistent with C.
(B) get_score (argument C ': estimation result)
Processing content: For each element of the estimation result, an evaluation value for each element is calculated.
(C) get_max (argument C ′: correction target, Score: evaluation value set of reliability of the estimation result, argument Rj: living situation rule, argument cnt: number of elements to be corrected for C ′)
Processing content: For the correction target C ′, a combination of living situations that satisfies Rj and maximizes the total evaluation value is acquired.
(D) replace (argument C: original estimation result, C ″: correction result)
Processing content: The correction result is applied to the original estimation result.

  In the source code shown in FIG. 7, in this embodiment, the rule Rj is R9 = {one person staying at home, bathing, cooking}, and C is {one person at home, bathing, cooking} It becomes. In get_common which is the first process, only the estimation result (element) related to the rule Rj is extracted from C which is the estimation result and is set as C ′. This is a process for excluding elements unrelated to the rule. However, in this embodiment, since Rj and C are the same set, C ′ = C.

  Next, the number of elements to be corrected cnt is initially set to 1, and one element (estimation result) is corrected at C ′. Here, an evaluation value is calculated for each set of various estimation results obtained by correcting one element of C ′, and a set having the maximum total evaluation value, which is the sum of the evaluation values, that is, a set having the highest evaluation is obtained. Next, when a consistent set is obtained, this set is used as a set of corrected estimation results (by performing the replacement process in FIG. 7), and the correction process is terminated.

  On the other hand, if a consistent set cannot be obtained, 1 is added to cnt, and the same calculation is performed for correcting two elements in C ′ to confirm whether a consistent set is obtained.

Here, a method of calculating an evaluation value for each element (estimation result) will be described. In this embodiment, the evaluation value score is a value obtained by multiplying the estimation accuracy score1 (0 ≦ score1 ≦ 1) of the estimation unit itself and the reliability score score2 (0 ≦ score2 ≦ 1) of each estimation result. The evaluation value score related to the reliability of the estimation unit for the living situation c is calculated as follows using the power consumption amount x _{t at the} time _t and the estimation condition of the living situation shown in FIG.
(1) score (x _t , c) = score1 (c) × score2 (x _t , c) where
score1 (c) ＝ | estimated result set 生活 correct answer life situation set | / | estimated result set |
score2 (x _t , c) = 1− (| x _t -average value of power consumption condition | / range of power consumption condition)
Among the above formulas, the average value of the power consumption condition is the average value of the upper and lower limits in the estimation condition related to the power consumption of the living situation shown in FIG. It can be the difference between the upper limit and the lower limit.

About score1 (c), it evaluates in the state where the living situation set used as a correct answer is known, and is derived. On the other hand, score2 (x _t , c) is obtained in the form of the following equations (2) and (3) for “during bathing” and “during cooking”, for example, when the estimation condition of FIG. 4 is used.
(2) score2 (x _t , bathing) = 1− (| x _t- (75 + 65) / 2 | / (75-65))
(3) score2 (x _t , bathing) = 1− (| x _t- (75 + 55) / 2 | / (75-55))

Furthermore, the total evaluation value total_score (x _t ), which is an evaluation index for the entire estimation result, is calculated as the sum of score (x _t , c) as in the following equation (4).
(4) total_score (x _t ) = Σscore (x _t , c)
Here, Σ is a summation about elements c belonging to the set C, where C is a set of living situations estimated to be true.

It should be noted that it is also possible to obtain the evaluation value score (x _t , c) simply by setting score (x _t , c) = score1 (c) without calculating score2 (x _t , c). However, when calculating the evaluation value score (x _t , c), the reliability score score2 (x _t , c) of the estimation result is used, so that the actual power consumption of the user who is the estimation target and the reliability of the estimation condition are trusted. It is possible to obtain an evaluation value with higher accuracy considering the characteristics.

  FIG. 8 is a table showing evaluation values for each estimation result in the present embodiment. Here, FIG. 8A shows the evaluation value of each estimation result at time 7:30 when the power consumption is 70 kWh, and FIG. 8B shows the evaluation value at time 19:00 when the power consumption is 73 kWh. The evaluation value of each estimation result is shown.

Using the evaluation values shown in FIG. 8A, the total evaluation values when only one element of C ′ at time 7:30 is corrected are expressed by the following equations (5) to (7), respectively.
(5) score (70, 1 person at home) + score (70, cooking) = 0.81 + 0.60 = 1.41
(6) score (70, 1 person staying at home) + score (70, taking a bath) = 0.81 + 0.35 = 1.16
(7) score (70, bathing) + score (70, cooking) = 0.35 + 0.60 = 0.95

From the above results, it can be seen that the combination that maximizes the total evaluation value total_score (x _t ), that is, the combination with the highest evaluation is score (70, one person at home) + score (70, cooking). As a result, at time 7:30, the set is corrected to a set of estimation results of “one person is at home”, “not taking a bath”, and “cooking”. Similarly, at time 19:30, the maximum value of the total evaluation value total_score (x _t ) is score (73,1 at home) + score (73, bathing) = 0.81 + 0.49 = 1.30. It is corrected to a set of estimation results of “at home”, “taking a bath”, and “not cooking”.

  Returning to FIG. 2, the estimation result correction unit 114 that has performed the correction processing described above outputs a set of corrected estimation results as a final estimation result.

  The evaluation value derivation process and the estimation result correction process described above are one example, and the living situation estimation unit of the present invention is naturally not limited to the above. For example, when not only the amount of power consumed by the main power but also the amount of power consumed by individual distribution systems or individual home appliances is obtained, it is also possible to use a living situation estimation unit that uses such amount of power. is there. In this case, the score2 calculation method described above changes. For example, when the power consumption amount of the lighting device is acquired and the operating status of the lighting is known, it is possible to perform a calculation considering the operating status of the lighting in addition to the main power in the determination of sleeping.

Further, the total evaluation value, which is an evaluation index of the entire estimation result, and further the estimation accuracy and reliability score are not limited to the definitions described above. For example, the evaluation value score can be defined in various forms as long as it is a monotonically decreasing function with respect to the difference between the power consumption amount xt at time _t and the average value of the estimation conditions related to the power consumption amount.

  As described above in detail, according to the living situation estimation apparatus, program and method of the present invention, a plurality of living situations are estimated based on the power consumption, and the consistency within the set of these estimation results is Judgment is made using living situation rules. Thereby, it is possible to obtain consistent estimation results for a plurality of living situations.

  It is also preferable that the living situation rule can be selected or changed according to user information. In this case, a more accurate combination of estimation results according to the user's actual situation can be obtained.

  Furthermore, in the living situation estimation apparatus, program and method of the present invention, when adding a living situation item to be estimated, a living situation estimation means corresponding to the item is added to the apparatus, and a rule is added to the “living situation rule”. Just add it to support the addition. As a result, even when it is desired to estimate a comprehensive life situation across multiple items, it can be easily applied.

  In addition, based on the information on the living situation of the user (household) estimated in this way as described above, appropriate support information about the user's power consumption, for example, power saving advice and household management advice, and further suitable for the living situation. It is also possible to provide product / service information. This also makes it possible to realize energy saving and marketing that is more suitable for the needs of users (households).

  Various changes, modifications, and omissions of the above-described various embodiments of the present invention can be easily made by those skilled in the art. The above description is merely an example, and is not intended to be restrictive. The invention is limited only as defined in the following claims and the equivalents thereto.

DESCRIPTION OF SYMBOLS 1 Living situation estimation apparatus 101 Communication interface part 102 Power consumption data storage part 111a 1st living situation estimation part 111b 2nd living situation estimation part 111n Nth living situation estimation part 112 Estimation result confirmation part 112a Living situation rule management part 112b User information management unit 112c Consistency determination unit 113 Evaluation value management unit 114 Estimation result correction unit 2 User information management device 3 Smart meter 4 HGW
5 carrier communication network

Claims (8)

An apparatus for estimating a user's living situation based on a user's power consumption,
A plurality of living situation estimation means for estimating a living situation for each of a plurality of living situation items to be estimated based on the data of the power consumption of the user;
An estimation result confirming means for judging consistency between the estimation results of the living situation in each of the plurality of living situation items based on the living situation rule that determines consistency between the plurality of living situations;
Evaluation value management means for holding an evaluation value calculated based on the estimation accuracy in each of the plurality of living situation estimation means;
For a set of estimation results determined to be inconsistent, based on an evaluation value relating to the estimation result of the set, one or more estimation results of the estimation result of the set are changed to be in a consistent state And calculating the total evaluation value, and comparing the calculated total evaluation values with each other, an estimation result correcting unit that sets the set of the estimated results after the change with the highest evaluation as a set of the corrected estimation results. A user life situation estimation device characterized by comprising:   The said evaluation value management means hold | maintains the evaluation value calculated based on the said estimation precision and the reliability score of the estimation result of the living condition by the said living condition estimation means. User life situation estimation device. The living state estimating means sets a condition relating to power consumption for the living situation item to be estimated, estimates a living situation based on whether the power consumption satisfies the condition,
The evaluation value management means calculates the reliability score using the power consumption amount in the estimation target time zone or time and the value that defines the condition used for the estimation, and for each estimation result of the living situation, Calculate the evaluation value based on the product of the estimation accuracy of the living situation estimation means and the calculated reliability score,
The estimation result correcting unit is configured to estimate one or a plurality of estimation results of the set of estimation results for the set of estimation results determined to be inconsistent based on the calculated evaluation value related to the estimation result of the set. When the result is changed to a consistent state, the total value of the evaluation values related to the unchanged estimation results is calculated as the total evaluation value, and the set of the estimation results after the change in which the total evaluation value becomes the maximum, The user life situation estimation apparatus according to claim 1 or 2, wherein the set is a set of corrected estimation results.   At least one of the plurality of living situation estimation means estimates a living situation for a living situation item related to the number of people staying in the household area of the user, and at least one of the remaining living situation estimation means The user life situation estimation apparatus according to any one of claims 1 to 3, wherein a life situation is estimated for a life situation item related to a living behavior of the user in the household area.   The said estimation result confirmation means selects or changes the living condition rule used in the case of judgment of consistency based on the user information which concerns on the said user's life, Either of Claim 1 to 4 characterized by the above-mentioned. The user life situation estimation apparatus according to item 1.   At least one of the plurality of life situation estimation means is not limited to the data on the total power consumption of the user, or instead of the data on the total power consumption, the power consumption by a specific device used by the user 6. The user life situation estimation apparatus according to claim 1, wherein the life situation is estimated based on the data. A life situation estimation program for causing a computer mounted on a device that estimates a user's life situation based on a user's power consumption to function,
A plurality of living situation estimation means for estimating a living situation for each of a plurality of living situation items to be estimated based on the data of the power consumption of the user;
An estimation result confirming means for judging consistency between the estimation results of the living situation in each of the plurality of living situation items based on the living situation rule that determines consistency between the plurality of living situations;
Evaluation value management means for holding an evaluation value calculated based on the estimation accuracy in each of the plurality of living situation estimation means;
For a set of estimation results determined to be inconsistent, based on an evaluation value relating to the estimation result of the set, one or more estimation results of the estimation result of the set are changed to be in a consistent state The computer is used as an estimation result correction unit that calculates the total evaluation value in the case of the above, and compares the calculated total evaluation values with each other to make the set of the estimated results after the change having the highest evaluation the set of the corrected estimation results. A living situation estimation program characterized by having a function. A life situation estimation method for estimating a user's life situation based on a user's power consumption,
A first step of estimating a living situation for each of a plurality of living situation items to be estimated based on the data of the power consumption of the user;
A second step of determining the consistency between the estimation results of the living situation in each of the plurality of living situation items based on the living situation rule that determines the consistency between the plurality of living situations;
A third step of holding an evaluation value calculated based on the estimation accuracy in each estimation in the first step;
For a set of estimation results determined to be inconsistent, based on an evaluation value relating to the estimation result of the set, one or more estimation results of the estimation result of the set are changed to be in a consistent state And calculating the total evaluation value in the case of the above, and comparing the calculated total evaluation values with each other, a fourth step of setting the set of the estimated result after the change with the highest evaluation as the set of the corrected estimation result, A life situation estimation method characterized by comprising:

Images