JP6964836B2 - Similarity calculation device, similarity calculation method, and program - Google Patents

Description

The present invention relates to a similarity calculation device, a similarity calculation method, and a program.

In so-called marriage hunting services such as matchmaking information provision services, it is important whether or not the registered members are compatible with each other. In the conventional marriage hunting service, registered members are asked to fill in information such as hobbies, tastes, income, and family structure on a questionnaire, and compatibility is estimated from the similarity of these conditions and introduced. In addition, in product sales services such as online shopping, services such as presenting products that are presumed to be of interest or interest to customers are provided based on the customer's search history and purchase history.

In the survey using the questionnaire in the marriage hunting service, only the subjective thoughts of the registered members could be grasped, and each information was fragmentary, and it was not possible to grasp the living conditions of the writer. On the other hand, in order to understand the compatibility, it is important to know whether or not the elements such as the rhythm and habits of life that the registered members themselves are not aware of match. Further, in the product sales service, it is not possible to present the recommended information based on the customer's life only by the product information based on the search history and the purchase history.

A living condition estimation device that consistently estimates a plurality of living conditions based on the amount of power consumption is known (Patent Document 1).
The living condition estimation device described in Patent Document 1 estimates the living condition for each of a plurality of living condition items to be estimated based on the data of the power consumption of the user. The living condition estimation device described in Patent Document 1 determines the consistency between the estimation results of the living conditions in each of the plurality of living condition items based on the living condition rules that negotiate the consistency between the living conditions.
The living condition estimation device described in Patent Document 1 determines a simple threshold value for power consumption and estimates whether the person is at home or not. The living condition estimation device described in Patent Document 1 divides power consumption data into predetermined times, converts them into feature vectors, and clusters them using a machine learning method such as a random forest method. Further, the living condition estimation device described in Patent Document 1 estimates the living condition based on the data of the power consumption of the specific device used by the user.

Japanese Unexamined Patent Publication No. 2015-153156

However, the living condition estimation device described in Patent Document 1 has a drawback that it cannot calculate the similarity of living patterns between consumers based on the amount of power consumption only by estimating the living condition.

The present invention has been made in view of the above points, and provides a similarity calculation device, a similarity calculation method, and a program capable of calculating the similarity of living patterns between consumers based on the amount of power consumption.

(1) The present invention has been made to solve the above-mentioned problems, and one aspect of the present invention is calculated from a load curve showing a power consumption amount for a predetermined period of a consumer and the load curve. A calculation unit that calculates the similarity between the consumers of at least one curve of the duration curve and the differential value curve obtained by differentiating the road curve based on the shape of the curve, and the calculation unit. Based on the calculated degree of similarity, the consumer classification unit that classifies the consumers who have similar shapes of the curve, the customer classified by the consumer classification unit, and the type of life pattern of the curve. It is a similarity calculation device including a life pattern determination unit that determines a specific life pattern that is similar among the consumers based on the associated features and the shape of the curve.

( 2 ) Further, in one aspect of the present invention, in the similarity calculation device, the feature is a feature of a part of the curve.

( 3 ) Further, in one aspect of the present invention, in the similarity calculation device, the calculation unit calculates the similarity for a specific period.

( 4 ) Further, one aspect of the present invention is a load curve indicating the power consumption of a consumer for a predetermined period, a duration curve calculated from the load curve, and a differential value obtained by differentiating the load curve. at least one of the similarity between the customer mutual curve and the first step of calculating, based on the shape of the curve, the curve on the basis of the similarity calculated by the first step of the curve The second process of classifying the consumers having similar shapes, the customer classified by the second process, the characteristics associated with the type of life pattern of the curve, and the curve. This is a similarity calculation method in which a computer executes a third process of determining a specific life pattern that is similar among the consumers based on the shape of the above.

( 5 ) Further, one aspect of the present invention is obtained by differentiating the load curve indicating the power consumption of the consumer for a predetermined period, the duration curve calculated from the load curve, and the load curve into a computer. is based on at least one similarity between the customer cross curve of the differential value curve and the first step of calculating, based on the shape of the curve, to the degree of similarity calculated by the first procedure The second procedure for classifying the consumers who have similar shapes of the curves, the consumers classified by the second procedure, and the features associated with the types of life patterns of the curves. , A program for executing a third procedure for determining a specific life pattern that is similar among the consumers based on the shape of the curve.

According to the present invention, it is possible to calculate the similarity of life patterns between consumers based on the amount of power consumption.

It is a figure which shows an example of the outline of the similarity calculation system which concerns on 1st Embodiment of this invention. It is a schematic block diagram which shows an example of the structure of the similarity calculation apparatus which concerns on this embodiment. It is a figure which shows an example of the load curve, the duration curve, and the differential value curve which concerns on this embodiment. It is a flowchart which shows an example of the similarity calculation processing of the similarity calculation apparatus which concerns on this embodiment. It is a figure which shows an example of the load curve of the electric energy which concerns on this embodiment. It is a figure which shows an example of the similarity degree of the road curve of the electric energy which concerns on this embodiment. It is a schematic block diagram which shows an example of the structure of the similarity calculation apparatus which concerns on 2nd Embodiment of this invention. It is a flowchart which shows an example of the determination process of the similarity calculation apparatus which concerns on this embodiment. It is a figure which shows an example of the classification result of the consumer which concerns on this embodiment. It is a figure which shows an example of the feature information which concerns on this embodiment. It is a figure which shows an example of the road curve and the duration curve of the monthly power consumption which concerns on this embodiment. It is a figure which shows an example of the load curve and the differential value curve of the daily power consumption which concerns on this embodiment. It is a figure which shows an example of the road curve and the duration curve of the monthly power consumption which concerns on this embodiment.

(First Embodiment)
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing an example of an outline of the similarity calculation system according to the present embodiment. The similarity calculation system of the present invention includes a similarity calculation device 1, a power consumption management server 2, a consumer management server 3, and a house 4.

The similarity calculation device 1 acquires the power consumption information transmitted from the house 4 via the power consumption management server 2. Here, the power consumption information is information indicating the power consumption for each predetermined time consumed in the house 4. The similarity calculation device 1 calculates each curve of the power consumption load curve, the duration curve, and the differential value curve based on the acquired power consumption information. The similarity calculation device 1 calculates the similarity of life between consumers based on the calculated shape of each curve. Here, the consumer is a resident of the house 4 that uses electric power in the house 4.
The road curve, duration curve, and differential value curve of power consumption are considered to reflect the lifestyle patterns of consumers.
The similarity calculation device 1 provides the calculated similarity of life between consumers to the consumer management server 3. The consumer management server 3 is, for example, a server of a business operator that operates a marriage hunting service or online shopping.

The amount of power consumed in the house 4 is measured by the measuring device 40 at predetermined time intervals. Here, the predetermined time is, for example, 30 minutes. The management device 41 acquires the value of the power consumption measured by the measuring device 40. The management device 41 generates power consumption information from the acquired power consumption value. The management device 41 transmits the generated power consumption information to the power consumption management server 2 via the home gateway 42. The consumer can make the management device 41 output the power consumption information to the terminal device and check the power consumption.
The measuring device 40 may generate the power consumption information without going through the management device 41 and transmit it to the power consumption management server 2 via the home gateway 42. When the measuring device 40 transmits the power consumption information to the power consumption management server 2 without going through the management device 41, the measuring device 40 has a communication function. Further, when the measuring device 40 transmits the power consumption information to the power consumption management server 2 without going through the management device 41, the management device 41 may not be provided in the house 4.
Further, the measuring device 40 may transmit the power consumption information to the power consumption management server 2 without going through either the management device 41 or the home gateway 42. When the measuring device 40 transmits the power consumption information to the power consumption management server 2 without going through either the management device 41 or the home gateway 42, the measuring device 40 has a communication function. When the measuring device 40 transmits the power consumption information to the power consumption management server 2 without going through either the management device 41 and the home gateway 42, the management device 41 and the home gateway 42 are not provided in the house 4. May be good. The measuring device 40 may transmit power consumption information to the power consumption management server 2 via a terminal device such as a personal computer or a smartphone.

FIG. 2 is a schematic block diagram showing an example of the configuration of the similarity calculation device 1 according to the present embodiment.
The similarity calculation device 1 includes a control unit 10 and a communication unit 11.
The control unit 10 includes a power consumption information acquisition unit 100, a curve generation unit 101, a calculation unit 102, and a presentation unit 103.

The power consumption information acquisition unit 100 acquires power consumption information from the power consumption management server 2 via the communication unit 11.
The curve generation unit 101 generates a load curve indicating the power consumption for a predetermined period of the consumer from the power consumption information acquired from the power consumption information acquisition unit 100. The curve generation unit 101 calculates a duration curve from the generated road curve. The curve generation unit 101 differentiates the generated load curve and calculates a differential value curve.
Here, each curve generated by the curve generation unit 101 will be described with reference to FIG.

FIG. 3 is a diagram showing an example of a load curve, a duration curve, and a differential value curve according to the present embodiment.
FIG. 3A is a diagram showing an example of a road curve. The road curve is a curve that shows the power consumption of a consumer for a predetermined period as a function of time. The curve generation unit 101 generates a load curve by interpolating the power consumption at 30-minute intervals for a period of one year using a smooth curve.

FIG. 3B is a diagram showing an example of a duration curve. Here, the duration curve is a curve calculated by arranging the values of the power consumption of the road curve in descending order, ignoring the time series.
FIG. 3C is a diagram showing an example of a differential value curve. Here, the differential value curve is a curve obtained by differentiating the load curve with respect to time.

Returning to FIG. 2, the description of the configuration of the similarity calculation device 1 will be continued.
The calculation unit 102 calculates the similarity between consumers of at least one of the load curve, the duration curve, and the differential value curve based on the shape of the curve.
The presentation unit 103 presents the degree of similarity between the consumers of the curve calculated by the calculation unit 102 to the consumer management server 3 via the communication unit 11.
The communication unit 11 transmits and receives various types of information via the wireless network. The communication unit 11 is hardware for performing communication via a wireless network.

The power consumption management server 2 manages power consumption information for each consumer. A consumer identifier is associated with each of the power consumptions.
The consumer management server 3 manages various information about the consumer. Each of the various types of information about the consumer is associated with the consumer's identifier. Here, the various information about the consumer is, for example, the personal information of the consumer.
The identifier of the consumer managed by the power consumption management server 2 and the identifier of the consumer managed by the consumer management server 3 are associated with each other.

FIG. 4 is a flowchart showing an example of the similarity calculation process of the similarity calculation device 1 according to the present embodiment.
(Step S100) The power consumption information acquisition unit 100 acquires power consumption information from the power consumption management server 2 via the communication unit 11. The power consumption information acquisition unit 100 outputs the acquired power consumption information to the curve generation unit 101.

(Step S101) The curve generation unit 101 acquires power consumption information from the power consumption information acquisition unit 100. The curve generation unit 101 generates a load curve for each of the consumers from the acquired power consumption information. The curve generation unit 101 calculates a duration curve for each of the consumers from the generated road curve. The curve generation unit 101 differentiates the generated load curve and calculates a differential value curve for each of the consumers.
The curve generation unit 101 outputs a load curve, a duration curve, and a differential value curve for each consumer to the calculation unit 102.
The calculation unit 102 may normalize each curve by dividing the value of the power consumption included in each curve by the maximum value of the power consumption included in each curve.

(Step S102) The calculation unit 102 acquires a load curve, a duration curve, and a differential value curve for each customer from the curve generation unit 101. The calculation unit 102 calculates the degree of similarity between the consumers of the acquired road curve based on the shape of the curve. The calculation unit 102 calculates the similarity between the consumers of the acquired duration curve based on the shape of the curve. The calculation unit 102 calculates the similarity between the consumers of the acquired differential value curve based on the shape of the curve. The calculation unit 102 outputs information indicating the degree of similarity of each calculated curve to the presentation unit 103.
Here, the similarity of the curves will be described with reference to FIGS. 5 and 6.

FIG. 5 is a diagram showing an example of a road curve of electric energy according to the present embodiment.
In FIG. 5A, as an example, a road curve LC5a, a road curve LC10b, a road curve LC10c, and a road curve LC10d for four consumers are drawn.
The degree of similarity between the road curve LC5a and the road curve LC5b is larger than, for example, the degree of similarity between the road curve LC5a and the road curve LC5c, or the degree of similarity between the road curve LC5a and the road curve LC5d.

In FIG. 5B, as an example, a duration curve DC5a, a duration curve DC5b, a duration curve DC5c, and a duration curve DC5d are drawn for four consumers.
The similarity between the duration curve DC5a, the duration curve DC5b, and the duration curve DC5c is larger than the similarity between the duration curve DC5a, the duration curve DC5b, and the duration curve DC5c with respect to the duration curve DC5d.

In FIG. 5C, as an example, the differential value curve BC5a, the differential value curve BC5b, the differential value curve BC5c, and the differential value curve BC5d for four consumers are drawn.
The similarity between the differential value curve BC5a and the differential value curve BC5b is larger than, for example, the similarity between the differential value curve BC5a and the differential value curve BC5c, or the similarity between the differential value curve BC5a and the differential value curve BC5d.

FIG. 6 is a diagram showing an example of the degree of similarity of the load curve of the electric energy according to the present embodiment.
As an example, the calculation unit 102 calculates the similarity of the load curves by using the cosine similarity between the values of the power consumption every 30 minutes of the two load curves for which the similarity is calculated at each time. The calculation unit 102 calculates the cosine similarity using a vector having a value of the power consumption every 30 minutes of the load curve as a component at each time. The calculation unit 102 similarly calculates the similarity for the duration curve and the differential value curve.
The calculation unit 102 may calculate the similarity of the curve by using another known similarity instead of the cosine similarity. For example, the calculation unit 102 may use the Pearson correlation coefficient.

The calculation unit 102 may calculate the similarity using at least one of the load curve, the duration curve, and the differential value curve.
The calculation unit 102 calculates the similarity of one of the load curve, the duration curve, and the differential value curve, and when the similarity is smaller than a predetermined value, that is, when the curves are not similar to each other, the other You may calculate the similarity of the curves of. For example, the calculation unit 102 may calculate the similarity of the load curves, and may calculate the similarity of the duration curve and the differential value curve when the curves are not similar to each other.

The calculation unit 102 may calculate the similarity of curves for a specific period. The calculation unit 102 may calculate the similarity of each curve for half a year, three months, or one month, for example. Further, the calculation unit 102 may calculate the similarity of each curve only on Saturday, Sunday, or a holiday, for example. Further, the calculation unit 102 may calculate the similarity of each curve only on weekdays. Further, the calculation unit 102 may calculate the similarity of each curve for each day of the week. Further, the calculation unit 102 may calculate the similarity of each curve separately for daytime and nighttime.

Returning to FIG. 4, the description of the similarity calculation process will be continued.
(Step S103) The presentation unit 103 acquires information indicating the similarity of each curve from the calculation unit 102. The presentation unit 103 presents the acquired information indicating the similarity of each curve to the consumer management server 3 via the communication unit 11.

As described above, the similarity calculation device 1 according to the present embodiment includes a calculation unit 102. The calculation unit 102 includes at least one of a load curve indicating the amount of power consumed by the consumer for a predetermined period, a duration curve calculated from the load curve, and a differential value curve obtained by differentiating the load curve. The degree of similarity between the consumers is calculated based on the shape of the curve.

With this configuration, the similarity calculation device 1 according to the present embodiment can calculate the similarity of the power consumption between consumers, so that the lifestyle patterns between consumers are similar based on the power consumption. The degree can be calculated.

(Second Embodiment)
Hereinafter, the second embodiment of the present invention will be described in detail with reference to the drawings.
In the first embodiment, the case where the similarity calculation device calculates the similarity between consumers of each curve of the power consumption has been described. In the present embodiment, a case where the similarity calculation device classifies consumers based on the similarity will be described.
The similarity calculation device according to the present embodiment is referred to as a similarity calculation device 1a.

FIG. 7 is a schematic block diagram showing an example of the configuration of the similarity calculation device 1a according to the present embodiment. Comparing the similarity calculation device 1a (FIG. 7) according to the present embodiment and the similarity calculation device 1 (FIG. 2) according to the first embodiment, the presence / absence of the control unit 10a and the storage unit 12a is different. However, the functions of the other components (communication unit 11) are the same as those of the first embodiment. The description of the same function as that of the first embodiment is omitted, and in the second embodiment, a part different from that of the first embodiment will be mainly described.

The similarity calculation device 1a includes a control unit 10a, a communication unit 11, and a storage unit 12a.
The control unit 10a includes a power consumption information acquisition unit 100, a curve generation unit 101, a calculation unit 102, a presentation unit 103, a consumer classification unit 104a, and a life pattern determination unit 105a.
The consumer classification unit 104a classifies consumers who have similar shapes of the curves based on the similarity calculated by the calculation unit 102.
The life pattern determination unit 105a is similar between the consumers classified by the consumer classification unit 104a, the characteristics associated with the type of life pattern of each curve, and the shape of each curve. Determine a particular life pattern.

The storage unit 12a stores the feature information 120a in the feature information table. The feature information 120a is information in which the feature of each curve calculated by the curve generation unit 101 is associated with the type of life pattern of the consumer. Details of the feature information 120a will be described later.

FIG. 8 is a flowchart showing an example of the determination process of the similarity calculation device 1a according to the present embodiment.
Since the processes of step S200, step S202, and step S205 are the same as the processes of step S100, step S102, and step S103 in FIG. 4, the description thereof will be omitted.

(Step S201) The curve generation unit 101 outputs a load curve, a duration curve, and a differential value curve for each consumer to the life pattern determination unit 105a. Other processing is the same as the processing of step S101 in FIG.

(Step S203) The consumer classification unit 104a acquires information indicating the similarity of each curve from the calculation unit 102. The consumer classification unit 104a classifies each curve having a similar shape into a cluster based on the acquired similarity of each curve. The consumer classification unit 104a classifies consumers by classifying each curve into clusters. The consumer classification unit 104a determines that the consumers corresponding to the curves included in the cluster including a predetermined number or more of the curves are the consumers whose living patterns are similar for each cluster.
The consumer classification unit 104a outputs the classification result of the consumer to the life pattern determination unit 105a and the presentation unit 103.
It is not necessary to specify in advance how many clusters the consumer classification unit 104a classifies the curve into. The consumer classification unit 104a classifies each curve into clusters by using a known clustering method such as multidimensional scaling or hierarchical clustering.
Here, the classification result of consumers will be described with reference to FIG.

FIG. 9 is a diagram showing an example of the classification results of consumers according to the present embodiment.
Loading curves, each cluster _{Ai (i = 1,2, ···,} N A: N A is the number of clusters load curve is classified) are classified into. Duration curves, each cluster _{Bi (i = 1,2, ···,} N B: N B is the number of clusters duration curve is classified) are classified into. Differential value curves, each cluster _{Ci (i = 1,2, ···,} N C: N C is the number of clusters differential value curve is classified) are classified into.
However, the number of curves included in each cluster is one or more.

Returning to FIG. 8, the description of the determination process of the similarity calculation device 1a will be continued.
(Step S204) The life pattern determination unit 105a acquires the classification result of the consumer from the consumer classification unit 104a. The life pattern determination unit 105a acquires a load curve, a duration curve, and a differential value curve for each consumer from the curve generation unit 101. The life pattern determination unit 105a acquires the feature information 120a from the storage unit 12a.

The life pattern determination unit 105a provides characteristic information for each curve feature indicated by the acquired feature information 120a, for the curve indicated by the type of curve to be determined, and for all curves included in the cluster indicated by the classification result of the consumer. It is determined whether or not the curve has the characteristics of the curve indicated by 120a. In the life pattern determination unit 105a, when all the curves included in the cluster have the characteristics of the curves shown by the feature information 120a, the life patterns shown by the feature information 120a are similar for the consumers corresponding to the curves included in the cluster. Judge that there is.
That is, the life pattern determination unit 105a is used between the consumers based on the consumers classified by the consumer classification unit 104a, the characteristics associated with the types of life patterns of each curve, and the shape of each curve. Determine specific life patterns that are similar.

The life pattern determination unit 105a associates the clusters indicated by the classification results of the consumers with the life patterns determined to be similar among the consumers. The life pattern determination unit 105a outputs the classification result of the consumer to which the life pattern is associated to the presentation unit 103.
Here, the feature information 120a will be described with reference to FIG.

FIG. 10 is a diagram showing an example of feature information 120a according to the present embodiment. The correspondence between the characteristics of the curve and the type of life pattern will be explained in order.

The integral value of the load curve in a specific period indicates the power consumption in a specific period. Here, the specific period is, for example, one year, half a year, three months, one month, one week, or one day. In addition, weekdays, holidays, days of the week, or day and night may be added as a specific period.
The integral value of the load curve in a specific period is associated with the power consumption tendency in the specific period. The fact that the integral values of the load curves in a specific period are close to each other indicates that the power consumption tendency in a specific period is similar among consumers.

The life pattern determination unit 105a calculates the integrated value of each load curve included in the cluster in a specific period for each cluster of the load curve indicated by the classification result of the consumer. The life pattern determination unit 105a calculates a representative value of the calculated integrated value. Here, the representative value is, for example, an average value or a median value. The life pattern determination unit 105a determines whether or not the error of the calculated integrated value with respect to the calculated representative value is equal to or less than a predetermined ratio for each load curve included in the cluster.
In the life pattern determination unit 105a, when the error of the calculated integrated value with respect to the calculated representative value is less than or equal to a predetermined ratio for all the load curves included in the cluster, the power consumption tendency in a specific period is different between consumers. Judge that they are similar between them.

The characteristic that the road curve is smaller than the predetermined value for several days indicates that the person is taking a long vacation and traveling. By looking at the range in which the load curve is smaller than a predetermined value for several days, the time, number of times, and period of long vacations can be known. The range in which the load curve is smaller than a predetermined value for several days is associated with the travel preference.

Here, a method in which the life pattern determination unit 105a determines the travel preference degree will be described with reference to FIG.
FIG. 11 is a diagram showing an example of a monthly power consumption load curve and duration curve according to the present embodiment.
Whether or not the life pattern determination unit 105a has a feature that the power consumption of each load curve included in the cluster of each load curve indicated by the classification result of the consumer continues for several days to be smaller than a predetermined value. Is determined. The life pattern determination unit 105a determines that the timing and number of long vacations taken are similar among consumers when the power consumption is smaller than a predetermined value and continues for several days.

The life pattern determination unit 105a consumes the duration curve DC11a included in the cluster, the load curve LC11a corresponding to the duration curve DC11b, and the road curve LC11b for each cluster of the duration curve indicated by the classification result of the consumer. It is determined whether or not the state in which the amount of electric power is smaller than a predetermined value has a characteristic of continuing for several days. When the life pattern determination unit 105a has a feature that the load curve LC11a and the road curve LC11b are smaller than a predetermined value for several days, the acquisition period of the long vacation is similar among the consumers. judge. Here, the period for taking a long vacation includes not only the case of one long vacation but also the case of the total of a plurality of long vacations. The load curve LC11a has been in a state smaller than a predetermined value for several days in July. The load curve LC11b has been in a state smaller than a predetermined value for several days in August.

It should be noted that the period of long vacation distinguishes between overseas travel and domestic travel, and the state where the road curve is smaller than the predetermined value is associated with the range that lasts for 4 days or more and the overseas travel, and the road curve is the predetermined value. For example, the range in which the smaller state lasts for 2 days or more and less than 4 days may be associated with the preference for domestic travel.

Returning to FIG. 10, the description of the feature information 120a will be continued.
The characteristic that the differential value curve has an upward peak at a specific time indicates that the consumer has woken up. The time when the differential value curve has an upward peak indicates the wake-up time. On the other hand, the characteristic that the differential value curve has a downward peak at a specific time indicates that the consumer has gone to bed. The time when the differential value curve has a downward peak at a specific time indicates the bedtime.
The characteristic that the differential value curve has an upward peak at a specific time and the characteristic that the differential value curve has a downward peak at a specific time are associated with the life rhythm.

Here, with reference to FIG. 12, a method in which the life pattern determination unit 105a determines the life rhythm will be described.
FIG. 12 is a diagram showing an example of a differential value curve including a load curve of daily power consumption according to the present embodiment.
The life pattern determination unit 105a determines whether the differential value curve has an upward or downward peak in the vicinity of a specific time for each differential value curve included in the cluster for each cluster of the differential value curve indicated by the classification result of the consumer. Judge whether or not. The differential value curve BC12 has an upward peak from 6 o'clock to 8 o'clock and a downward peak from 0 o'clock to 2 o'clock. When the life pattern determination unit 105a determines that the differential value curve has an upward peak near a specific time, the life pattern determination unit 105a determines as the wake-up time the time when the curve starts to rise on the load curve LC12 corresponding to the differential value curve BC12. do. When the life pattern determination unit 105a determines that the differential value curve has a downward peak at a specific time, the life pattern determination unit 105a determines the time at which the curve starts to descend on the load curve LC12 corresponding to the differential value curve BC12 as the bedtime.
The life pattern determination unit 105a determines from the determined wake-up time and bedtime that the life rhythms are similar among the consumers.

Returning to FIG. 10, the description of the feature information 120a will be continued.
The characteristic that the differential value curve has an upward peak during the time of supper preparation indicates that supper preparation has begun at home. Here, the time zone for preparing dinner is, for example, from 17:00 to 19:00. The characteristic that the differential value curve has an upward peak in the time zone for preparing dinner is associated with the ratio of self-catering for dinner. A low percentage of self-catering for dinner means a high preference for eating out.

The life pattern determination unit 105a determines whether or not each of the differential value curve clusters indicated by the classification result of the consumer has an upward peak in the time zone for preparing dinner for each differential value curve included in the cluster. .. From the determination result, the lifestyle pattern determination unit 105a calculates the ratio of the differential value curve having an upward peak in the time zone for preparing dinner as the ratio of self-catering for dinner. The lifestyle pattern determination unit 105a determines that the eating out preferences are similar among the key households based on the calculated self-catering ratio of the dinner.

The feature that the load curve continues to be smaller than a predetermined value for several hours during the night time indicates that the consumer is taking a bath because it is considered that the electric device is not used during the bath. Here, the night time zone is, for example, from 18:00 to 21:00.

The life pattern determination unit 105a sets a time zone in which a state smaller than a predetermined value continues for several hours in the night time zone for each load curve included in the cluster for each load curve cluster indicated by the classification result of the consumer. judge. The length of the determined time zone is associated with the bathing preference.

The feature that the load curve continues to be larger than a predetermined value for several hours indicates that an air conditioner such as an air conditioner is used. The feature that the load curve continues to be larger than a predetermined value for several hours is associated with the resistance to temperature changes.

The feature that the road curve is smaller than a predetermined value on a specific day of the week such as Saturday or Sunday, or a holiday indicates the number of times of going out or the time of going out on a specific day of the week or a holiday. The feature that the road curve is smaller than the average value on weekdays on a specific day of the week or a holiday is associated with the outdoor preference.

Here, with reference to FIG. 13, a method in which the life pattern determination unit 105a determines the outdoor preference degree will be described.
FIG. 13 is a diagram showing an example of a monthly power consumption load curve and duration curve according to the present embodiment.
The life pattern determination unit 105a compares the load curve to a predetermined value on a specific day of the week or a holiday with respect to the load curve corresponding to each duration curve included in each cluster of the duration curve indicated by the classification result of the consumer. Determine if there is a small day. When the life pattern determination unit 105a determines that there is a day when the load curve is smaller than a predetermined value on a specific day of the week or a holiday, the life pattern determination unit 105a determines the number of outings and the outing time from the load curve. The life pattern determination unit 105a determines that the outdoor preference is similar among consumers based on the determined number of outings and outing time.
The road curve LC13a corresponding to the duration curve DC13a is smaller than a predetermined value on Saturdays and Sundays 6, 7, 13, 14, 20, 21, 27 and 28. The road curve LC13b corresponding to the duration curve DC13b is smaller than the predetermined value on Wednesday and Sunday on 3, 7, 10, 14, 17, 21, 24 and 28 days.

Returning to FIG. 9, the description of the determination process of the similarity calculation device 1a will be continued.
(Step S205) The presentation unit 103 acquires the classification result of the consumer to which the life pattern is associated from the life pattern determination unit 105a. The presentation unit 103 presents the acquired classification result of the customer to the customer management server 3 via the communication unit 11.

As described above, the similarity calculation device 1a according to the present embodiment includes a consumer classification unit 104a and a life pattern determination unit 105a.
The consumer classification unit 104a classifies consumers having similar curve shapes (road curve, duration curve, and differential value curve) based on the similarity calculated by the calculation unit 102.
The life pattern determination unit 105a is similar between the consumers classified by the consumer classification unit 104a, the characteristics associated with the type of life pattern of the curve, and the shape of the curve. Determine a particular life pattern.

With this configuration, the similarity calculation device 1a according to the present embodiment can classify the consumers based on the similarity of the curve indicating the power consumption, so that the consumers can be classified based on the living pattern.
Further, since the similarity calculation device 1a according to the present embodiment can determine a specific life pattern that is similar between consumers, it is possible to present a life pattern that explains the classification of consumers.

In addition, the characteristics associated with the types of life patterns of the curve are the characteristics of a part of the curve.
With this configuration, the similarity calculation device 1a according to the present embodiment can determine a life pattern in which a feature appears only in a part of the curve. Therefore, as a life pattern for explaining the classification of consumers, the feature is only in a part of the curve. It is possible to present the life patterns that appear.

In addition, the calculation unit 102 calculates the degree of similarity for a specific period.
With this configuration, the similarity calculation device 1a according to the present embodiment can determine a life pattern in which the characteristics appear only in a specific period. Therefore, a life pattern in which the characteristics appear only in a specific period as a life pattern for explaining the classification of consumers. Patterns can be presented.

A part of the similarity calculation devices 1 and 1a in the above-described embodiment, for example, the control unit 10 and the control unit 10a may be realized by a computer. In that case, the program for realizing this control function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read by the computer system and executed. The "computer system" referred to here is a computer system built in the similarity calculation devices 1 and 1a, and includes hardware such as an OS and peripheral devices. Further, the "computer-readable recording medium" refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, or a CD-ROM, or a storage device such as a hard disk built in a computer system. Furthermore, a "computer-readable recording medium" is a medium that dynamically holds a program for a short period of time, such as a communication line when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In that case, a program may be held for a certain period of time, such as a volatile memory inside a computer system serving as a server or a client. Further, the above-mentioned program may be a program for realizing a part of the above-mentioned functions, and may be a program for realizing the above-mentioned functions in combination with a program already recorded in the computer system.
Further, a part or all of the similarity calculation devices 1 and 1a in the above-described embodiment may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each functional block of the similarity calculation devices 1 and 1a may be individually made into a processor, or a part or all of them may be integrated into a processor. Further, the method of making an integrated circuit is not limited to the LSI, and may be realized by a dedicated circuit or a general-purpose processor. Further, when an integrated circuit technology that replaces an LSI appears due to advances in semiconductor technology, an integrated circuit based on this technology may be used.

Although one embodiment of the present invention has been described in detail with reference to the drawings, the specific configuration is not limited to the above, and various design changes and the like are made without departing from the gist of the present invention. It is possible to do.

1, 1a ... similarity calculation device, 2 ... power consumption management server, 3 ... consumer management server, 4 ... housing, 10, 10a ... control unit, 11 ... communication unit, 12a ... storage unit, 40 ... measurement device, 41 ... Management device, 42 ... Home gateway, 100 ... Power consumption information acquisition unit, 101 ... Curve generation unit, 102 ... Calculation unit, 103 ... Presentation unit, 104a ... Consumer classification unit, 105a ... Life pattern determination unit, 120a … Feature information

Claims (5)

The consumer of at least one curve of a load curve showing the power consumption of the consumer for a predetermined period, a duration curve calculated from the load curve, and a differential value curve obtained by differentiating the load curve. A calculation unit that calculates the degree of similarity between each other based on the shape of the curve ,
A consumer classification unit that classifies the consumers whose curve shapes are similar based on the similarity calculated by the calculation unit, and a consumer classification unit.
Based on the characteristics associated with the consumer classified by the consumer classification unit, the type of life pattern of the curve, and the shape of the curve, specific lives that are similar to each other among the consumers. Life pattern judgment unit that judges patterns and
A similarity calculation device comprising. The features and the similarity calculation device according to claim 1, which is a feature of some of the curves. The similarity calculation device according to claim 1 or 2 , wherein the calculation unit calculates the similarity for a specific period. The consumer of at least one curve of a load curve showing the electric energy consumption of the consumer for a predetermined period, a duration curve calculated from the load curve, and a differential value curve obtained by differentiating the load curve. The first process of calculating the similarity between each other based on the shape of the curve, and
A second process of classifying the consumers who have similar curve shapes based on the similarity calculated by the first process, and a second process.
Specific specific similarities between the consumers based on the consumer classified by the second process, the characteristics associated with the type of life pattern of the curve, and the shape of the curve. The third process of determining life patterns and
A computer-executed similarity calculation method. On the computer
The consumer of at least one curve of a load curve showing the electric energy consumption of the consumer for a predetermined period, a duration curve calculated from the load curve, and a differential value curve obtained by differentiating the load curve. The first procedure for calculating the similarity between each other based on the shape of the curve, and
A second procedure for classifying the consumers who have similar curve shapes based on the similarity calculated by the first procedure, and
Specific specific similarities between the consumers based on the consumer classified by the second procedure, the characteristics associated with the type of life pattern of the curve, and the shape of the curve. The third procedure to determine the life pattern and
A program to execute.

Images