JP7621804B2 - Planning support device and method - Google Patents

Description

The present invention relates to a planning support device and method, and is suitable for use in a planning support device that creates capital investment plans for power infrastructure operators such as power companies and power distribution companies.

In recent years, the environment surrounding the energy business has been changing significantly, with changes in medium- to long-term electricity demand resulting from factors such as an increase in renewable energy and a decline in electricity demand due to population decline, changes in the relationship between new entities and existing electricity infrastructure operators, and the opening of an electricity trading market. As a result, electricity infrastructure operators are being called upon to respond to these changes in the environment.

In light of these circumstances, electricity infrastructure operators are being forced to switch to a new electricity network in order to improve supply stability, environmental compatibility, and efficiency while maximizing the use of existing facilities. Against this background, there is a need to create an environment in which the electricity demand of electricity consumers (hereafter simply referred to as consumers) can be properly understood and consumers' electricity demand can be reduced as necessary. There is also a growing demand for new businesses that provide new services to consumers whose electricity demand has been understood.

As background technology for this application, Patent Document 1 discloses a method of generating multiple power consumption models from first power consumption information that stores various information about the power consumption of each consumer that has a smart meter installed, and then identifying a power consumption model that satisfies a similarity criterion from among the multiple models for second power consumption information that stores various information about each consumer that has only a power meter installed but no smart meter installed, and estimating the amount of power consumption.

Patent Document 2 also discloses a power distribution system state estimation system that uses equipment information on the power distribution line and measurement information from measuring instruments at some nodes to estimate the state (active power, reactive power, voltage, etc.) of loads and power sources installed at nodes where no measuring instruments are installed.

Furthermore, Patent Document 3 discloses a method for obtaining a representative load pattern from the long-term power usage data accumulated in the past by collecting the actual power usage data of each power consumer for a specified period of time, calculating the feature parameters that best represent the shape of all load patterns for the load patterns of multiple consumers for the number of days during that period, and forming clusters based on the similarity of the feature parameters.

JP 2015-106952 A JP 2010-263754 A JP 2008-15921 A

However, the inventions disclosed in Patent Documents 1 to 3 are unable to accurately grasp the power demand characteristics of consumers under uncertain conditions. This means that power infrastructure operators must create investment plans that allow for a significant margin of error when making capital investments, which can lead to the problem of being forced to make excessive capital investments that take stability into account.

The present invention has been made in consideration of the above points, and aims to propose a planning support device and method that can curb excessive capital investment and improve the profitability and business viability of power infrastructure projects.

In order to solve such problems, in the present invention, a planning support device that supports the development of investment plans for power distribution facilities is provided with a consumer information complementation unit that complements values of items whose values are unknown in consumer information, which is predetermined information related to the power demand of each consumer that has been acquired in advance, a consumer load pattern calculation unit that calculates or estimates a load pattern of each consumer based on the complemented consumer information for each consumer , a load calculation unit that calculates the load at each point in the power distribution system based on the calculated or estimated load pattern of each consumer, and a power demand sensitivity analysis unit that calculates the sensitivity of the power demand to influencing factors that affect the power demand for each consumer , wherein the influencing factors are temperature and/or time, and the consumer load pattern calculation unit corrects the load pattern of each consumer based on the sensitivity of the power demand of each consumer to the influencing factors previously calculated by the power demand sensitivity analysis unit.

In addition, in the present invention, a planning support method executed by a planning support device that supports the planning of investment plans for power distribution facilities includes a first step of complementing values of items whose values are unknown among consumer information, which is predetermined information related to the power demand of each consumer that has been acquired in advance, a second step of calculating or estimating a load pattern of each consumer based on the complemented consumer information for each consumer, a third step of calculating a load at each point in the power distribution system based on the calculated or estimated load pattern of each consumer, and a fourth step of calculating, for each consumer, the sensitivity of the power demand to influencing factors that affect the power demand , wherein the influencing factors are temperature and/or time, and in the second step, the load pattern of each consumer is corrected based on the sensitivity of the power demand of each consumer to the influencing factors calculated in the fourth step previously executed.

The planning support device and method of the present invention can grasp the power demand characteristics of consumers with high accuracy. This reduces the need for power infrastructure operators to make excessive capital investments that take stability into account.

The present invention provides a planning support device and method that can curb excessive capital investment and improve the profitability and business viability of power infrastructure projects.

1 is a block diagram showing a configuration of a planning support device according to an embodiment of the present invention; FIG. 2 is a block diagram showing a logical configuration of the planning support device. 1 is a diagram showing an example of the configuration of a consumer information database. 1 is a table showing an example of the configuration of a consumer load database. 1 is a diagram showing an example of the configuration of a plan database. 13 is a flowchart showing a processing procedure of a consumer information complementation process. 13 is a flowchart showing the processing steps of a consumer clustering process. 4 is a block diagram showing an example of the configuration of a consumer load pattern calculation unit. FIG. 4 is a block diagram showing an example of the configuration of a planning unit; FIG. 13 is a flowchart showing a processing procedure of an electric power demand sensitivity analysis process. 11 is a diagram illustrating a power demand sensitivity analysis process. FIG. 11 is a diagram illustrating a power demand sensitivity analysis process. FIG. FIG. 13 is a block diagram illustrating a configuration example of a load calculation and planning unit according to a second embodiment.

One embodiment of the present invention will be described in detail below with reference to the drawings.

(1) First embodiment (1-1) Configuration of the planning support device according to this embodiment In Fig. 1, the planning support device according to this embodiment is generally designated by 1. This planning support device 1 is configured from a general-purpose computer device equipped with a CPU (Central Processing Unit) 2, a memory 3, and an auxiliary storage device 4.

The CPU 2 is a processor that controls the operation of the entire planning support device 1. The memory 3 is composed of, for example, a volatile semiconductor memory, and is used as a work memory for the CPU 2. When the planning support device 1 is started up or when necessary, the consumer information complementation program 5, consumer load pattern calculation program 6, load calculation/planning program 7, and power demand sensitivity analysis program 8, which are described below, are read from the auxiliary storage device 4 and stored in the memory 3.

The auxiliary storage device 4 is composed of a large-capacity non-volatile storage device such as a hard disk device, SSD (Solid State Drive) or flash memory, and is used to store programs and necessary information for a long period of time. The consumer information database 9, consumer load database 10, plan database 11 and weather information database 12, which will be described later, are stored and maintained in this auxiliary storage device 4.

(1-2) Planning Support Function Next, a planning support function installed in the planning support device 1 will be described. This planning support function is a function that supports the planning of investment plans for power distribution facilities by power infrastructure operators. In this embodiment, a demand response execution plan (hereinafter, this is called a demand response plan) that limits the power demand of all consumers in each time period (hereinafter, this is simply called a time period) in 30-minute units at each point of the power distribution system to the amount of power that can be supplied by the power infrastructure operator is created, thereby supporting the planning of investment plans for power distribution facilities that can limit excessive investment.

In this case, a "point" may refer to a single point identified by latitude and longitude, or may refer to the entire area within a map divided into a 10 km x 10 km mesh.

As a means for realizing such a planning support function, the planning support device 1 is provided with a consumer information database 9, a consumer load database 10, a planning database 11, and a weather information database 12, as well as a consumer information complementation unit 20, a consumer load pattern calculation unit 21, a load calculation/planning unit 22, and a power demand sensitivity analysis unit 23, as shown in FIG. 2.

The consumer information database 9 is a database in which various information about each consumer who has concluded a power supply contract with a power infrastructure business operator is registered, and as shown in FIG. 3, has a table-like structure with a consumer ID column 9A, an address column 9B, a contract information column 9C, a contract capacity column 9D, a business type column 9E, a number of residents column 9F, an owned home appliance column 9G, and a daytime at-home rate column 9H. In the consumer information database 9, one row corresponds to one consumer. Note that, hereinafter, the various information about a consumer is collectively referred to as the consumer information of that consumer.

The consumer ID column 9A stores an identifier (consumer ID) unique to the corresponding consumer that is assigned to that consumer, and the address column 9B stores the consumer's address (the location where the consumer consumes electricity). The contract information column 9C stores the electricity supply plan that the corresponding consumer has concluded with the electricity infrastructure business as contract information, and the contract capacity column 9D stores the maximum capacity in the electricity supply contract that the consumer has concluded with the electricity infrastructure business.

The business type column 9E stores the business type of the corresponding consumer (general household, shop, office, factory, etc.), and the number of residents column 9F stores the number of residents living at the address stored in the address column 9B. Furthermore, the owned appliances column 9G stores all the types of appliances owned by the corresponding consumer and their power consumption per unit time, and the daytime at-home rate column 9H stores the daytime at-home rate for that consumer.

Therefore, in the example of Figure 3, for example, a consumer with a consumer ID of "0x3333" has an address of "△ Prefecture, △ City, △ Town, △ Chome, △ Number △", has a contract with the power infrastructure provider for "Pay-As-You-Go B" (40A), is an "ordinary household" with "4" residents, owns one "refrigerator," one "washing machine," and one "air conditioner" each with a power consumption of "x**kW," and is at home "50% of the time."

The information registered in the consumer information database 9 is information obtained from each consumer when the power infrastructure operator enters into a power supply contract with the consumer, or through questionnaires conducted at other times. For unknown information, the corresponding column in the consumer information database 9 is left blank.

It should also be noted that accurate information such as the number of residents and the rate of people at home during the day is not always available. In this case, the daytime rate of people at home can be estimated to a reasonably close extent by apportioning the same contracted capacity and number of residents based on the value (%) from a statistical survey on the rate of people at home every hour during the day from the "National Life Statistics" published by the Japan Broadcasting Corporation (NHK). Even if the number of residents is unknown, the daytime rate of people at home can be estimated from the contracted capacity alone.

The consumer load database 10 is a database that stores the actual power load (hereinafter referred to as power consumption, as appropriate) of each consumer for each time period, obtained from the smart meters installed in each consumer, and as shown in FIG. 4, has a table-like structure with a consumer ID column 10A, a date column 10B, and multiple actual column 10C. In the consumer load database 10, one row corresponds to one consumer.

The consumer ID column 10A stores the consumer ID of the corresponding consumer, and the date column 10B stores the date on which the results for that row were obtained. The results columns 10C are set up to correspond to each 30-minute time period starting from "00:00", and each results column 10C stores the measured power consumption of the corresponding consumer in the corresponding time period.

Therefore, in the example of Figure 4, the power consumption on "January 1, 2020" of a consumer with a consumer ID of "0x2222" is shown to be "1 kWh" during the "00:00-00:30" time period, "0.5 kWh" during the "00:30-01:00" time period, "0.4 kWh" during the "01:00-01:30" time period, "0.3 kWh" during the "01:30-02:00" time period, "0.3 kWh" during the "02:00-02:30" time period, ... and "0.8 kWh" during the "23:30-00:00" time period.

The plan database 11 is a database used to hold the above-mentioned demand response plans that have been created, and as shown in FIG. 5, has a table-like structure with a plan ID column 11A, a pair of location columns 11BA and 11BB, a target period column 11C, a demand response amount column 11D, and a service information call column 11E. In the plan database, one row corresponds to one demand response plan.

The plan ID column 11A stores an identifier (plan ID) that is unique to the corresponding demand response plan and is assigned to that demand response plan. The location columns 11BA and 11BB are provided to correspond to the locations at both ends of a section of a power distribution system where power consumption should be reduced by demand response, and the identifier (location ID) of the location at one end of the section is stored in one location column 11BA, and the location ID of the location at the other end of the section is stored in the other location column 11BB.

Furthermore, the target period column 11C stores the execution period of the corresponding demand response plan, and the demand response amount column 11D stores the amount of power consumption reduction in the demand response plan for that plan. Furthermore, the service information call column stores an identifier (service information) of a service to be called to the corresponding consumer when the corresponding demand response plan is executed. Note that an example of such a "service" could be a service that provides points to consumers who cooperate with demand response.

Therefore, in the example of Figure 5, the demand response plan assigned the plan ID "0x00000011" reduces power consumption by a total of "10 kWh" through demand response for consumers connected to the section of the distribution system from point "0xXXXX0011" to point "0xXXXX0022" during the period "YYYYMMDD HHMMSS_1 to YYYYMMDD HHMMSS_2", and the service information for the service to be called to the corresponding consumers at this time is "DM1".

The weather information database 12 is a database that stores past weather information such as temperature, humidity, and wind conditions at each point in the power distribution system for each time period, and predicted future weather information such as temperature, humidity, and wind conditions at each point in the power distribution system for each time period.

On the other hand, the consumer information complementation unit 20 is a functional unit that is realized by the CPU 2 (Figure 1) executing the consumer information complementation program 5 (Figure 1) stored in the memory 3 (Figure 1). The consumer information complementation unit 20 executes a process to complement missing information in the consumer information of each consumer based on the consumer information of each consumer registered in the consumer information database 9, and outputs the consumer information of each consumer after complementation to the consumer load pattern calculation unit 21.

The consumer load pattern calculation unit 21 is a functional unit that is realized by the CPU 2 executing the consumer load pattern calculation program 6 (FIG. 1) stored in the memory 3. The consumer load pattern calculation unit 21 executes a clustering process to group consumers with similar load patterns (power consumption patterns) based on the consumer information of each consumer that has been supplemented by the consumer information supplementation unit 20, the power consumption by time period of each consumer registered in the consumer load database 10, and meteorological information such as temperature, humidity, and wind conditions by time period in the past that is registered in the meteorological information database 12.

The consumer load pattern calculation unit 21 also estimates the load pattern of each consumer (e.g., a new consumer) whose load pattern is unknown based on the results of the clustering process. The consumer load pattern calculation unit 21 then outputs the load pattern for each consumer grouped by the clustering process and the load pattern for each new consumer obtained by the estimation to the load calculation/planning unit 22.

The load calculation and planning unit 22 is a functional unit that is realized by the CPU 2 executing the load calculation and planning program 7 (FIG. 1) stored in the memory 3. The load calculation and planning unit 22 calculates the load at each point on the distribution system in each time period based on the load pattern of each consumer provided by the consumer load pattern calculation unit 21 and meteorological information such as temperature, humidity, and wind conditions for each future time period registered in the meteorological information database 12.

The load calculation/planning unit 22 also creates a demand response plan based on the calculated load at each point on the power distribution system for each time period, determining whether or not to implement demand response in each future time period and, if so, how to implement it. The load calculation/planning unit 22 registers the created demand response plan in the plan database 11 and outputs it to the power demand sensitivity analysis unit 23.

The power demand sensitivity analysis unit 23 is a functional unit that is realized when the CPU 2 executes the power demand sensitivity analysis program 8 (FIG. 1) stored in the memory 3. The power demand sensitivity analysis unit 23 calculates the sensitivity of the power demand of each consumer (hereinafter referred to as power demand sensitivity) to each factor that affects the power demand, such as temperature and time (hereinafter referred to as influencing factors), and notifies the consumer load pattern calculation unit 21 of the calculated power demand sensitivity of each consumer to each influencing factor.

The consumer load pattern calculation unit 21 corrects the load pattern for each consumer based on the power demand sensitivity to each influencing factor for each consumer provided by the power demand sensitivity analysis unit. This allows the consumer load pattern calculation unit 21 to calculate a more accurate load pattern for each consumer.

(1-3) Details of Each Functional Unit (1-3-1) Details of Consumer Information Complement Unit Fig. 6 shows specific processing content of consumer information complementation processing executed by the consumer information complementation unit 20. When a user issues an instruction to create a demand response plan (hereinafter, this instruction is referred to as a demand response plan creation instruction), the consumer information complementation unit 20 executes consumer information complementation processing to complement values of missing items of each consumer information registered in the consumer information database 9 (in the example of Fig. 2, "address", "contract information", "contract capacity", "business type", "number of residents", "owned home appliances", and/or "daytime at-home rate") based on the consumer information of other consumers, according to the processing procedure shown in Fig. 6.

In the following, it is assumed that consumer information for each consumer is obtained by conducting a survey in which the answers to each question are multiple choice, such as when concluding an electricity supply contract. In this case, when formulating the questions in the survey, in addition to questions directly related to electricity consumption, redundant questions not directly related to electricity consumption, such as favorite sports, favorite book genres, and travel preferences, are also prepared. By performing consumer clustering as described below based on the consumer's answers to each question, it is possible to obtain the effect of making it easier to classify consumers, albeit partially.

Returning to the explanation of Figure 6, when the demand response plan creation instruction is given, the consumer information complementation unit 20 starts the consumer information complementation process shown in Figure 6, and first executes a consumer clustering process (S1) to cluster each consumer into multiple groups (hereinafter referred to as consumer groups) with similar situations and preferences based on the values of each item of the consumer information of each consumer stored in the consumer information database 9 and the responses to the questionnaire.

Next, the consumer information complementation unit 20 judges whether or not the processing from step S4 onwards has been completed for all consumer groups generated in step S1 (S2). If the consumer information complementation unit 20 obtains a negative result in this judgment, it selects one consumer group for which processing from step S4 onwards has not been completed (S3).

Next, the consumer information complementation unit 20 calculates the characteristic vector of each consumer in the consumer group selected in step S3 (hereinafter referred to as the selected consumer group) (S4). Specifically, for each consumer, the consumer information complementation unit 20 generates a multi-dimensional vector whose vector components are each item of the consumer information of that consumer registered in the consumer information database 9 and the consumer's answer to each question in the questionnaire, and calculates the characteristic vector of that consumer by normalizing each vector component to the range of 0 to 1.

Next, the consumer information complementation unit 20 generates a classification tree for the selected consumer group based on the characteristic vector of the consumer information of each consumer in the selected consumer group calculated as described above (S5). Note that any existing method can be widely applied as a method for generating the classification tree at this time.

After that, for each group of consumers (hereinafter referred to as consumer small groups) in the selected consumer group classified by the classification tree generated in step S5, if there is an item whose value is unknown in the consumer information of each consumer belonging to that consumer small group, the consumer information complementing unit 20 complements the value using the maximum likelihood estimation method (S6). For example, if the value of an item in the consumer information of a certain consumer is unknown, the consumer information complementing unit 20 estimates the average or median value of that item in the consumer information of each consumer classified in the same consumer small group as the value of the unknown item and complements it. At this time, the consumer information complementing unit 20 assigns information such as a confidence interval to the value.

When the consumer information completion unit 20 has completed completing the values of items with unknown values in the consumer information of each consumer belonging to the selected consumer group in the manner described above, it returns to step S2 and determines whether or not the processing from step S4 onwards has been performed for all consumer groups (S2).

If the consumer information complementation unit 20 obtains a negative result in this determination, it proceeds to step S3, and then repeats the processing of steps S2 to S6 while sequentially switching the consumer group selected in step S3 to the consumer group that has not been processed in steps S4 and onward.

Then, when the consumer information completion unit 20 obtains a positive result in step S2 by completing completion of unknown items of the consumer information of each consumer for all consumer groups generated in step S1, it ends the consumer information completion process.

The specific processing contents of the consumer clustering process executed by the consumer information complementation unit 20 in step S1 of the consumer information complementation process are shown in FIG. 7.

When the consumer information complementation unit 20 proceeds to step S1 of the consumer information complementation process, it starts this consumer clustering process and first calculates the feature vectors of all consumers (S10) in the same manner as in step S4 of the consumer information complementation process.

Next, based on the survey results of consumers who answered all questions in the survey (hereinafter, these consumers are referred to as fully-responsive consumers), the consumer information complementation unit 20 clusters the feature vectors of these fully-responsive consumers into several groups using clustering techniques such as the k-means method (S11).

Next, for each consumer who did not answer all the questions in the questionnaire (hereinafter referred to as a non-fully answered consumer), the consumer information complementation unit 20 generates multiple feature vectors for that consumer with combination patterns of all options by assigning an arbitrary option to each unanswered question, and extracts combination patterns from among these generated feature vectors that have a reliability of 90% or higher (S12).

Furthermore, the consumer information complementation unit 20 classifies the feature vectors of all the consumers who have responded and the feature vectors extracted in step S12 for each of the consumers who have not responded into a plurality of groups by clustering using the k-means method or the like (S13). Each group created in this way corresponds to the consumer group described above.

Then, the consumer information complementation unit 20 judges whether the clustering has resulted in a rational grouping based on the number of classifications (number of consumer groups) and the bias of the elements of the feature vectors of each classified consumer (S14).

If the consumer information complementing unit 20 obtains a negative result in this judgment, it updates the values of each component of the characteristic vector of each consumer by the steepest descent method so as to reduce each vector component of the characteristic vector of each consumer in proportion to 1/(Euclidean distance)^2 (S15). The consumer information complementing unit 20 then returns to step S13, and thereafter executes the processes from step S13 onwards in the same manner as described above until a positive result is obtained in step S14. By doing this, sparse vector information such as questionnaires can be further aggregated.

Then, when the consumer information complementation unit 20 obtains a positive result in step S14 because it has finished classifying each consumer into a rational consumer group, it ends the consumer clustering process.

(1-3-2) Details of the consumer load pattern calculation unit Fig. 8 shows a specific configuration of the consumer load pattern calculation unit 21. As shown in Fig. 8, the consumer load pattern calculation unit 21 is configured to include a load pattern learning unit 21A and a load pattern estimation unit 21B.

The load pattern learning unit 21A clusters each consumer into multiple groups of consumers with similar load patterns (hereinafter, these are referred to as consumer groups with similar load patterns) based on the consumer information of each consumer after completion provided by the consumer information completion unit 20, the past power consumption of these consumers for each time period stored in the consumer load database 10, and the weather information for these time periods stored in the weather information database 12.

The load pattern learning unit 21A also classifies each consumer belonging to the similar load pattern consumer group into a plurality of smaller groups by creating a decision tree for each consumer group with a similar load pattern, using the address, contract information, contract capacity, business type, number of residents, and daytime at-home rate based on the consumer information of each consumer supplemented by the consumer information supplementing unit 20. The load pattern learning unit 21A then outputs the classification results for each consumer obtained in this manner, together with the load pattern of each group, to the load pattern estimation unit 21B as the learning results of the consumer load patterns.

The load pattern estimation unit 21B estimates the load patterns of consumers with unknown load patterns based on the load pattern learning results provided by the load pattern learning unit 21A and the consumer information of each consumer whose load pattern is unknown among the consumers whose consumer information is registered in the consumer information database 9 (e.g., a new consumer, hereinafter referred to as a consumer with unknown load pattern).

Specifically, the load pattern estimation unit 21B matches a consumer whose load pattern is known and whose load pattern is similar to the consumer's information based on the consumer's information about the consumer whose load pattern is unknown, and estimates the load pattern of the consumer whose load pattern is known as the load pattern of the consumer whose load pattern is unknown.

The load pattern estimation unit 21B then outputs the load pattern of each consumer whose load pattern is unknown obtained by such estimation and the load pattern of each consumer provided by the load pattern learning unit 21A to the load calculation/planning unit 22.

The load pattern estimation unit 21B also updates the load patterns of these consumers as necessary based on the load pattern learning results provided by the load pattern learning unit 21A, the consumer information of each consumer whose consumer information is registered in the consumer information database 9, and the formulas KPI1 and KPI2 provided by the power demand sensitivity analysis unit 23 (described below), and outputs the update results to the load calculation and planning unit 22.

(1-3-3) Details of the Load Calculation and Planning Unit Fig. 9 shows a specific configuration of the load calculation and planning unit 22. As shown in Fig. 9, the load calculation and planning unit 22 includes a load calculation unit 22A and a DR planning unit 22B.

The load calculation unit 22A calculates the load for each future time period at each point in the power distribution system based on the load pattern of each consumer provided by the consumer load pattern calculation unit 21 and meteorological information such as temperature, humidity, and wind conditions for each future time period registered in the meteorological information database 12. The load calculation unit 22A also calculates the available active power (P) and reactive power (Q) at each point based on the calculated load at each point.

Specifically, the load calculation unit 22A first calculates the load (hereinafter referred to as the total load) as the sum of the power demand at each point in the power distribution system at each time slice. The load calculation unit 22A also calculates the difference between the total load and the maximum load allowed for the power distribution system at each point as the load margin at that point. The load calculation unit 22A then uses the power factor of the supplied power to calculate the active power (P) and reactive power (Q) margins for each point.

The load calculation unit 22A then outputs the surplus amounts of active power (P) and reactive power (Q) at each point in the power distribution system calculated in this manner to the DR planning unit 22B.

The DR planning unit 22B plans a demand response plan, such as whether to implement a demand response or not, and how to implement the demand response if it is implemented, based on the available amount of active power (P) and reactive power (Q) at each point of the power distribution system provided by the load calculation unit 22A. The DR planning unit 22B then registers the planned demand response plan in the plan database 11, and outputs the planned demand response plan to the power demand sensitivity analysis unit 23.

(1-3-4) Details of the Power Demand Sensitivity Analysis Unit Fig. 10 shows specific processing contents of the power demand sensitivity analysis process executed by the power demand sensitivity analysis unit 23. The power demand sensitivity analysis unit 23 obtains a formula KPI1 for predicting power demand and a formula KPI2 for predicting the amount of power that can be reduced by demand response according to the processing procedure shown in Fig. 10.

In practice, when the power demand sensitivity analysis unit 23 receives a demand response plan from the load calculation and planning unit 22, it starts the power demand sensitivity analysis process shown in Figure 10, and first obtains the past power demand (power consumption) of each consumer for each time period from the consumer load database 10 (Figure 4), obtains weather information for these time periods from the weather information database 12 (Figure 1), and further obtains the values of each item of consumer information from the consumer information database (Figure 3) (S20).

Then, the power demand sensitivity analysis unit 23 detects the sensitivity of the power demand to each of the influencing factors such as the temperature and time for each consumer (S21). Specifically, as shown in FIG. 11 for example, the power demand sensitivity analysis unit 23 plots the relationship between the temperature and the power consumption of the corresponding consumer relative to the temperature, and calculates the slope of the straight line obtained by linear approximation. This slope is the sensitivity of the power demand of the corresponding consumer to the temperature. The power demand sensitivity analysis unit 23 similarly calculates the sensitivity of the power demand to each influencing factor for each consumer.

Next, the power demand sensitivity analysis unit 23 predicts the power demand (power consumption) of each consumer for each future time period based on the sensitivity of the power demand of each consumer to each influencing factor calculated in step S21 (S22). For example, the power demand sensitivity analysis unit 23 calculates the predicted value of the future power demand of each consumer for each time period by multiplying the predicted temperature for that time period by the temperature sensitivity calculated in step S21, and adding the intercept value of the corresponding line in FIG. 11 to the multiplication result. Similarly, the power demand sensitivity analysis unit 23 calculates the predicted value of the power demand of each consumer for each influencing factor.

Then, for each consumer, the power demand sensitivity analysis unit 23 determines the degree of influence of each influencing factor on each item of the consumer information of that consumer by ensemble learning of the power demand of that consumer in the past time period acquired in step S20, and sets a weight for each value of each item of the consumer information of that consumer based on the determination results (S23).

Then, based on the processing results of step S23, the power demand sensitivity analysis unit 23 calculates a formula (hereinafter referred to as the power consumption prediction formula KPI1) for each consumer to predict the amount of power consumption for each time period in the future (S24).

Then, the power demand sensitivity analysis unit 23 detects the sensitivity of the power demand to each influencing factor for each consumer, similar to step S21 (S25). In addition, the power demand sensitivity analysis unit 23 extracts characteristics for each influencing factor and for each time period by clustering analysis or the like, based on the sensitivity of the power demand to each influencing factor for each consumer detected in step S25 (S26).

For example, the power demand sensitivity analysis unit 23 performs clustering to classify consumers into multiple groups based on temperature sensitivity. As shown in FIG. 12, for each group obtained by this clustering, the power demand sensitivity analysis unit 23 partially differentiates the relational equation f between power consumption and temperature sensitivity in that group, and calculates the slope of the temperature sensitivity for each time period at the temperature predicted for that time period as the temperature characteristic for that time period. The same process is performed for other influencing factors, and characteristics are extracted for each group and time period.

Next, the electricity demand sensitivity analysis unit 23 calculates the probability f'(T) of a consumer responding to a demand response for each influencing factor and for each time period (hereinafter, this is called the DR induction rate) for that influencing factor (S27).

で与えられるシグモイド関数ｆ（Ｔ）を利用する。シグモイド関数は、変曲点（図１２の気温Ｔ_０の点）における傾きを数値で与えると形状が一意に決定できるため、扱いやすい特徴をもつ。 Specifically, the power demand sensitivity analysis unit 23 assigns a probability density function corresponding to the calculated gradient of the temperature sensitivity for each temperature to each group and time period. In this embodiment, the following formula is used as the probability density function.

The sigmoid function f(T) is used, which is given by: The sigmoid function is easy to handle because its shape can be uniquely determined by giving a numerical value for the slope at the inflection point (the point of temperature _T0 in FIG. 12).

Then, the power demand sensitivity analysis unit 23 calculates the DR induction rate f'(T) of the influencing factor for each time period by partially differentiating the characteristics of the influencing factor (sigmoid function f(T)) for each group for each influencing factor and for each time period.

Furthermore, the power demand sensitivity analysis unit 23 calculates the formula KPI2 for each consumer to calculate the amount of reduction in power consumption due to demand response, based on the future demand response plan for each time period provided by the load calculation/planning unit 22 and the DR induction rate f'(T) calculated in step S27 (S28).

Then, the power demand sensitivity analysis unit 23 outputs the formula KPI1 calculated in step S24 and the formula KPI2 calculated in step S28 to the load pattern estimation unit 21B (Figure 8) of the consumer load pattern calculation unit 21 (Figure 8) (S29), and then ends this power demand sensitivity analysis process.

(1-4) Effects of this embodiment As described above, the planning support device 1 of this embodiment complements unknown consumer information of each consumer, and calculates a load pattern of each consumer based on the complemented consumer information of each consumer. The planning support device 1 also calculates the power demand at each point in the power distribution system based on the calculated load pattern of each consumer, and creates a demand response plan based on the calculated power demand. At this time, the planning support device 1 calculates the sensitivity of the power demand to each influencing factor for each consumer, and corrects the load pattern of each consumer based on the calculated sensitivity for each consumer.

Therefore, according to the planning support device 1 of this embodiment, the load pattern of each consumer can be calculated with high accuracy, so that the load at each point in the power distribution system can be calculated with high accuracy, and accordingly, a highly accurate demand response plan can be created. Therefore, according to this planning support device 1, by suppressing the power consumption of each consumer according to the demand response point plan that has been created, it is possible to supply power to each consumer without making excessive investments in power distribution facilities, and therefore it is possible to suppress investments in excessive power distribution facilities and improve the profitability and business viability of power infrastructure operators.

1 and 2, a planning support device according to a second embodiment is generally designated by 30. This planning support device 30 is configured in the same manner as the planning support device according to the first embodiment, except that the functions of a planning program 31 (FIG. 1) and a load calculation and planning unit 32 (FIG. 2) realized by the CPU 2 (FIG. 1) executing the planning program 31 are different from the load calculation and planning program 7 and the load calculation and planning unit 22 of the first embodiment, respectively.

Figure 13 shows the configuration of the load calculation and planning unit 32 of this embodiment. As shown in this Figure 13, the load calculation and planning unit 32 of this embodiment is composed of a load calculation unit 32A and a power distribution equipment operation planning unit 32B. Of these, the function of the load calculation unit 32A is the same as that of the load calculation unit 22A of the first embodiment, so a description thereof will be omitted here.

On the other hand, the power distribution equipment operation plan creation unit 32B creates a power distribution equipment operation plan (hereinafter, referred to as a power distribution equipment operation plan) that specifies how to expand or reduce the power distribution equipment and how to operate the power distribution equipment.

In other words, while the DR planning unit 22B in the first embodiment creates a plan to suppress the power consumption of each consumer in accordance with the current power distribution equipment, the power distribution equipment operation planning unit 32B in the present embodiment creates a plan to expand or reduce the power distribution equipment and operate the power distribution equipment in accordance with the predicted power consumption of the consumer.

The power distribution equipment operation plan creation unit 32B then stores the power distribution equipment operation plan thus created in the plan database 11 and outputs it to the power demand sensitivity analysis unit 23.

According to the planning support device 30 of this embodiment, a power distribution facility operation plan can be created based on highly accurate predictions of power consumption by consumers, so that power can be supplied to each consumer without investing in excessive power distribution facilities. This in turn reduces investment in excessive power distribution facilities, improving the profitability and business viability of power infrastructure operators.

(3) Other embodiments In the above-mentioned first and second embodiments, the planning support device 1, 30 is configured as shown in Fig. 1, but the present invention is not limited to this, and for example, the consumer information supplement program 5, consumer load pattern calculation program 6, load calculation/planning program 7, and power demand sensitivity analysis program 8, and some or all of the consumer information database 9, consumer load database 10, plan database 11, and weather information database 12 may be distributed and arranged in different computer devices connected to each other via a network. Even in this way, the same effects as those of the first and second embodiments can be obtained.

In the first and second embodiments described above, the seven items of consumer information are "address," "contract information," "contract capacity," "business type," "number of residents," "owned appliances," and "daytime at-home rate." However, the present invention is not limited to this, and items other than these seven may be added or replaced as such items.

Furthermore, in the above-mentioned first and second embodiments, the consumer load pattern calculation unit 21 is configured as shown in FIG. 8, and the load calculation and planning units 22, 32 are configured as shown in FIG. 9 or FIG. 13. However, the present invention is not limited to this, and various other configurations can be widely applied as the configurations of the consumer load pattern calculation unit 21 and the load calculation and planning units 22, 32.

The present invention can be applied to a planning support device that supports the creation of investment plans for power distribution facilities.

1, 30... planning support device, 2... CPU, 5... consumer information supplement program, 6... consumer load pattern calculation program, 7, 31... load calculation and planning program, 8... power demand sensitivity analysis program, 9... consumer information database, 10... consumer load database, 11... planning database, 12... weather information database, 20... consumer information supplement unit, 21... consumer load pattern calculation unit, 21A... load pattern learning unit, 21B... load pattern estimation unit, 22, 32... load calculation and planning unit, 22A, 32A... load calculation unit, 22B... DR planning unit, 23... power demand sensitivity analysis unit, 32B... power distribution equipment operation planning unit.

Claims (12)

A planning support device that supports the creation of investment plans for power distribution facilities,
a consumer information complementing unit that complements values of items whose values are unknown among consumer information, which is predetermined information related to the power demand of each consumer acquired in advance;
A consumer load pattern calculation unit that calculates or estimates a load pattern of each of the consumers based on the supplemented consumer information for each of the consumers ;
a load calculation unit that calculates a load at each point of the power distribution system based on the calculated or estimated load pattern of each of the consumers;
and an electric power demand sensitivity analysis unit that calculates, for each of the consumers, a sensitivity of the electric power demand to an influencing factor that affects the electric power demand,
The influencing factors are temperature and/or time of day;
The consumer load pattern calculation unit,
a power demand sensitivity analysis unit that calculates a sensitivity of the power demand of each of the consumers to the influencing factors and that corrects the load pattern of each of the consumers based on the sensitivity of the power demand of each of the consumers to the influencing factors, the sensitivity being calculated in advance by the power demand sensitivity analysis unit. The consumer information complementing unit,
The planning support device according to claim 1 , wherein a value of the item of the consumer information, the value of which is unknown, is complemented based on a value of the item of another consumer. The consumer information complementing unit,
The planning support device according to claim 2 , wherein the value of the item for which the value of the consumer is unknown is complemented by a maximum likelihood estimation method based on the values of the item for other consumers. The consumer load pattern calculation unit,
The planning support device according to any one of claims 1 to 3, characterized in that each of the consumers whose past power consumption for each time period is known is classified based on the load pattern, and for a consumer whose past power consumption for each time period is unknown, the load pattern of the consumer is estimated to be the load pattern corresponding to a group to which other consumers whose information related to electricity demand has similar values belong. The power demand sensitivity analysis unit is
Calculating a characteristic of the sensitivity of the electricity demand to the impact factors;
determining a probability density function based on the calculation results;
Calculating a probability that the consumer will respond to a demand response using the determined probability density function;
calculating a formula for each consumer for determining an amount of reduction in the power demand due to the demand response based on the calculated probability;
The consumer load pattern calculation unit,
The planning support device according to claim 1 , further comprising: a power demand sensitivity analysis unit that calculates the load pattern of each of the customers by using the formula for each of the customers calculated by the power demand sensitivity analysis unit. 6. The planning support device according to claim 5, wherein the probability density function is a sigmoid function. A planning support method executed by a planning support device that supports investment planning for power distribution equipment, comprising:
A first step of complementing values of items whose values are unknown among consumer information, which is predetermined information related to the power demand of each consumer acquired in advance;
A second step of calculating or estimating a load pattern of each of the consumers based on the supplemented consumer information for each of the consumers ;
A third step of calculating a load at each point of the power distribution system based on the calculated or estimated load pattern of each of the consumers;
and a fourth step of calculating, for each of the consumers, a sensitivity of the power demand to an influencing factor that affects the power demand,
The influencing factors are temperature and/or time of day;
In the second step,
and correcting the load pattern of each of the customers based on the sensitivity of the power demand of each of the customers to the influencing factors calculated in the fourth step executed in advance. In the first step,
The planning support method according to claim 7 , wherein the value of the item of the consumer information, the value of which is unknown, is complemented based on the value of the item of other consumers. In the first step,
The planning support method according to claim 8, characterized in that the value of the item for which the value of the consumer is unknown is complemented by a maximum likelihood estimation method based on the values of the item for other consumers. In the second step,
The planning support method according to any one of claims 7 to 9, characterized in that each of the consumers whose past power consumption for each time period is known is classified based on the load pattern, and for a consumer whose past power consumption for each time period is unknown, the load pattern of the consumer is estimated to be the load pattern corresponding to a group to which other consumers whose information related to electricity demand has similar values belong. In the fourth step,
Calculating a characteristic of the sensitivity of the electricity demand to the impact factors;
determining a probability density function based on the calculation results;
Calculating a probability that the consumer will respond to a demand response using the determined probability density function;
calculating a formula for each consumer for determining an amount of reduction in the power demand due to the demand response based on the calculated probability;
In the second step,
The method for supporting planning according to any one of claims 7 to 10, further comprising: correcting the load pattern of each of the consumers by using the formula for each of the consumers calculated in the fourth step executed in advance. The method for supporting planning according to claim 11, wherein the probability density function is a sigmoid function.

Images