Power management device and power management method

Abstract

To provide a power management device or a power management method by which a facility that can appropriately make a contribution to the power demand-supply balance of a power system can be selected.SOLUTION: A power management device includes a management unit that manages two or more facilities connected to a power system, and a control unit that selects a combination of target facilities for making a contribution to the power demand-supply balance of the power system. The control unit estimates power probability distributions regarding the two or more facilities, and selects a combination of the target facilities in such a way that a combination of the power probability distributions exhibits a unimodal characteristic.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a power management apparatus and a power management method. [Background technology]

[0002] BACKGROUND ART In recent years, a technology (for example, a VPP (Virtual Power Plant)) that uses a power storage device as a distributed power source in order to maintain the balance between power supply and demand in a power grid has become known. In such cases, it is extremely important to predict the amount of power that can be adjusted by using a power storage device (for example, the amount of power that can be reduced from the power demand).

[0003] For example, as a method for predicting reducible power, a technology has been proposed that selects a combination of facilities that contributes to adjusting the balance of power supply and demand so as to reduce the variation (variance) of the expected value of reducible power for each facility relative to the expected value of the total reducible power for two or more facilities (e.g., Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2015-104137 Summary of the Invention [Problem to be solved by the invention]

[0005] In this regard, facilities that can contribute to adjusting the balance between power supply and demand may include a mixture of facilities with large-scale power storage devices and facilities with small-scale power storage devices. In such a case, the probability distribution of the reducible power of each facility may become multi-peaked.

[0006] As a result of extensive investigation, the inventors have focused on the above-mentioned points and found that the following problems exist.

[0007] Generally, assuming a case where the probability distribution of the reducible power of each facility is multi-peaked, simply using the expected value of the total reducible power may result in situations where it is not possible to appropriately select a facility with little variation in the expected value of the total reducible power, or where the expected value of the total reducible power may deviate from the actual reducible power.

[0008] Therefore, the present invention has been made to solve the above-mentioned problems, and aims to provide a power management device or a power management method that makes it possible to select facilities that can appropriately contribute to the power supply and demand balance in a power grid. [Means for solving the problem]

[0009] One aspect of the disclosure is a power management device that includes a management unit that manages two or more facilities connected to a power grid, and a control unit that selects a combination of target facilities that contributes to the power supply and demand balance of the power grid, wherein the control unit predicts a probability distribution of power for the two or more facilities and selects the combination of target facilities so that the combination of the probability distribution is unimodal.

[0010] One aspect of the disclosure is a power management method comprising step A of managing two or more facilities connected to a power grid, and step B of selecting a combination of target facilities that contributes to the power supply and demand balance of the power grid, wherein step B includes a step of predicting a probability distribution of power for the two or more facilities, and a step of selecting the combination of target facilities so that the combination of the probability distribution is unimodal. [Effects of the Invention]

[0011] According to the present invention, it is possible to provide a power management device or a power management method that makes it possible to select facilities that can appropriately contribute to the balance of power supply and demand in a power grid. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 1 is a diagram showing a power management system 1 according to an embodiment. [Figure 2] FIG. 2 is a diagram showing a facility 100 according to the embodiment. [Figure 3] FIG. 3 is a diagram illustrating a power management server 200 according to the embodiment. [Figure 4] FIG. 4 is a diagram for explaining a problem associated with the embodiment. [Figure 5] FIG. 5 is a diagram for explaining selection of a combination of target facilities according to the embodiment. [Figure 6] FIG. 6 is a diagram for explaining selection of a combination of target facilities according to the embodiment. [Figure 7] FIG. 7 is a diagram illustrating a power management method according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, embodiments will be described with reference to the drawings. In the following description of the drawings, the same or similar parts are denoted by the same or similar reference numerals. However, the drawings are schematic.

[0014] [Embodiment] (Power Management System) A power management system according to an embodiment will be described below. The power management system may be simply referred to as a power system.

[0015] As shown in FIG. 1, the power management system 1 includes two or more facilities 100 and a power management server 200.

[0016] Here, the facility 100 and the power management server 200 are configured to be able to communicate with each other via a network 11. The network 11 may include the Internet, a dedicated line such as a VPN (Virtual Private Network), or a mobile communication network.

[0017] The facility 100 is connected to the power grid 12 and may receive power from the power grid 12 or may supply power to the power grid 12. Power from the power grid 12 to the facility 100 may be referred to as forward flow power, purchased power, or demand power. Power from the facility 100 to the power grid 12 may be referred to as reverse flow power or sold power. In FIG. 1 , facilities 100A to 100C are illustrated as examples of the facility 100.

[0018] Although not particularly limited, facility 100 may be a facility such as a residence, a facility such as a store, or a facility such as an office. Facility 100 may also be an apartment building including two or more residences. Facility 100 may also be a complex including at least two or more of the following facilities: a residence, a store, and an office. Details of facility 100 will be described later (see FIG. 2).

[0019] The power management server 200 is managed by a business operator that manages power related to the power system 12. The business operator may be a power generation business operator, a power transmission and distribution business operator, or a retail business operator. The business operator may be a resource aggregator (hereinafter referred to as RA) or an aggregation coordinator (AC) that manages the RA. The RA may be a business operator that adjusts the power supply and demand balance of the power system 12. The adjustment of the power supply and demand balance may include a transaction in which reduced power of the demand power (forward flow power) of the facility 100 is exchanged for value (hereinafter referred to as negawatt trading). The adjustment of the power supply and demand balance may include a transaction in which increased power of reverse flow power is exchanged for value. In a VPP, the RA may be a business operator such as a power generation business operator, a power transmission and distribution business operator, or a retail business operator.

[0020] In the embodiment, communication between the power management server 200 and the EMS 160 is performed according to a first protocol. On the other hand, communication between the EMS 160 and the distributed power source (the solar cell device 110, the power storage device 120, or the fuel cell device 130) is performed according to a second protocol different from the first protocol. For example, the first protocol may be a protocol conforming to Open ADR (Automated Demand Response) or a unique dedicated protocol. For example, the second protocol may be a protocol conforming to ECHONET Lite (registered trademark), SEP (Smart Energy Profile) 2.0, KNX, or a unique dedicated protocol. Note that the first protocol and the second protocol may be different from each other; for example, even if both are unique dedicated protocols, they may be protocols created according to different rules. However, the first protocol and the second protocol may be protocols created according to the same rules.

[0021] (facility) A facility according to an embodiment will be described below. As shown in Fig. 2, the facility 100 includes a solar cell device 110, a power storage device 120, a fuel cell device 130, a load device 140, and an EMS (Energy Management System) 160. The facility 100 may also include a measuring device 190.

[0022] The solar cell device 110 is a distributed power source that generates electricity in response to light such as sunlight. For example, the solar cell device 110 is configured by a PCS (Power Conditioning System) and a solar panel. Here, installation may mean connecting the solar cell device 110 to the power grid 12.

[0023] The power storage device 120 is a distributed power source that charges and discharges power. For example, the power storage device 120 is configured by a PCS and a power storage cell. Here, "installed" may mean that the power storage device 120 is connected to the power grid 12.

[0024] The fuel cell device 130 is a distributed power source that generates electricity using fuel. For example, the fuel cell device 130 is composed of a PCS and a fuel cell. Here, "installed" may mean that the fuel cell device 130 and the power grid 12 are connected.

[0025] For example, the fuel cell device 130 may be a solid oxide fuel cell (SOFC; Solid Oxide Fuel Cell), a polymer electrolyte fuel cell (PEFC; Polymer Electrolyte Fuel Cell), a phosphoric acid fuel cell (PAFC; Phosphoric Acid Fuel Cell), or a molten carbonate fuel cell (MCFC; Molten Carbonate Fuel Cell).

[0026] The load devices 140 are devices that consume power. For example, the load devices 140 may include air conditioners, lighting devices, video devices, audio devices, refrigerators, washing machines, personal computers, and the like.

[0027] The EMS 160 manages the power related to the facility 100. The EMS 160 may control the solar cell device 110, the power storage device 120, the fuel cell device 130, and the load devices 140. In the embodiment, the EMS 160 is illustrated as an apparatus that receives control commands from the power management server 200, but such an apparatus may also be referred to as a gateway or simply as a control unit. To distinguish the EMS 160 from the power management server 200, the EMS 160 may also be referred to as a local EMS (LES), a home EMS (HEMS), or a VPP controller.

[0028] The measuring device 190 measures forward flow power from the power system 12 to the facility 100. The measuring device 190 may also measure reverse flow power from the facility 100 to the power system 12. For example, the measuring device 190 may be a smart meter belonging to a power company. The measuring device 190 may transmit an information element indicating a measurement result (an integrated value of forward flow power or reverse flow power) at a first interval (e.g., 30 minutes) to the EMS 160 at the first interval. The measuring device 190 may also transmit an information element indicating a measurement result at a second interval (e.g., 1 minute) that is shorter than the first interval to the EMS 160.

[0029] (Power management server) The power management server according to the embodiment will be described below. In the embodiment, a case where the power management server 200 is an RA will be described as an example. As shown in FIG. 3 , the power management server 200 includes a communication unit 210, a management unit 220, and a control unit 230.

[0030] The communication unit 210 is configured by a communication module. The communication module may be a wireless communication module conforming to standards such as IEEE802.11a / b / g / n / ac / ax, ZigBee, Wi-SUN, LTE, 5G, or 6G, or may be a wired communication module conforming to standards such as IEEE802.3.

[0031] For example, the communication unit 210 may communicate with the facility 100. The communication unit 210 may receive information indicating the power demand of the facility 100. The communication unit 210 may receive an operation plan for each distributed power source (the solar cell device 110, the power storage device 120, and the fuel cell device 130) from the facility 100. The communication unit 210 may receive an operation plan for the load device 140 from the facility 100. The communication unit 210 may transmit a control command for each distributed power source (the solar cell device 110, the power storage device 120, and the fuel cell device 130) to the facility 100. The communication unit 210 may transmit a control command for the load device 140 to the facility 100. The communication unit 210 may transmit a control command for controlling the forward flow power or reverse flow power of the facility to the facility 100.

[0032] The management unit 220 is configured by a storage medium such as a hard disk drive (HDD), a solid state drive (SSD), or a nonvolatile memory.

[0033] The management unit 220 manages information relating to two or more facilities 100. For example, the information relating to the facility 100 includes the type of distributed power source (solar cell device 110, power storage device 120, or fuel cell device 130) provided in the facility 100, the specifications of the distributed power source (solar cell device 110, power storage device 120, or fuel cell device 130) provided in the facility 100, etc. The specifications may include the rated power generation of the solar cell device 110, the rated charging power of the power storage device 120, the rated discharging power of the power storage device 120, and the rated output power of the fuel cell device 130. The specifications may also include the rated capacity of the power storage device 120, the maximum charging and discharging power, etc.

[0034] The management unit 220 may manage the power demand of the facility 100. The management unit 220 may manage the operation plans of the distributed power sources (the solar cell device 110, the power storage device 120, and the fuel cell device 130). The management unit 220 may manage the operation plans of the load devices 140. The management unit 220 may manage the operation mode of the EMS 160.

[0035] In the embodiment, the management unit 220 constitutes a management unit that manages two or more facilities 100 connected to the power grid 12 .

[0036] The control unit 230 may include at least one processor. The at least one processor may be configured by a single integrated circuit (IC), or may be configured by multiple circuits (such as integrated circuits and / or discrete circuits) that are communicatively connected.

[0037] In the embodiment, the control unit 230 configures a control unit that selects a combination of target facilities that contributes to the balance of power supply and demand in the power grid 12. The control unit 230 predicts a probability distribution of power for two or more facilities 100, and selects a combination of target facilities such that the combination of the probability distributions is unimodal.

[0038] In the embodiment, a case will be described in which the power related to two or more facilities 100 is reduced power from the power demand of two or more facilities 100. For example, the control unit 230 executes the following operation.

[0039] First, control unit 230 predicts a probability distribution of reduced power of power demand. Although not particularly limited, the probability distribution of reduced power of power demand may be predicted based on the predicted result of power demand and the predicted result of adjustable range of power storage device 120. An existing method may be used as a method for predicting the probability distribution.

[0040] For example, the control unit 230 may predict power demand by learning past power demand and factor data. The learning may be machine learning, such as AI (Artificial Intelligence). The factor data may include weather data, time data, etc. The weather data may include temperature, humidity, solar radiation, wind speed, air pressure, etc. The time data may include year, season, month, date, day of the week, etc.

[0041] For example, the control unit 230 may predict the adjustable range of the power storage device 120 based on an operation plan for the power storage device 120. The operation plan for the power storage device 120 may be formulated by the facility 100 (for example, the EMS 160).

[0042] Second, the control unit 230 selects a combination of target facilities so that the combination of probability distributions of reduced power demand is unimodal. The control unit 230 may select a combination of target facilities so that the most frequent value of the combination of probability distributions falls within a certain range (hereinafter referred to as the success determination range) based on the target reduced power. The control unit 230 may select a combination of target facilities so that the target reduced power falls within the power range that can be adjusted by the distributed power sources installed in the target facilities (for example, the adjustable range of the power storage device 120). The method of selecting a combination of target facilities will be described later (see FIGS. 5 and 6).

[0043] The control message specifying the target power reduction may be transmitted from the AC to the RA.

[0044] Third, the control unit 230 evaluates whether to participate in a control message (e.g., a DR; Demand Response request) received from the AC. The control unit 230 evaluates whether to participate in a DR request based on a combination of probability distributions corresponding to a combination of target facilities. For example, the control unit 230 may evaluate to participate in the DR request if the mode of the combination of probability distributions falls within a success determination range based on the target reduced power, and may evaluate not to participate in the DR request if the mode of the combination of probability distributions does not fall within the success determination range based on the target reduced power. The control unit 230 may evaluate to participate in the DR request if the target reduced power falls within the adjustable range of the power storage device 120, and may evaluate not to participate in the DR request if the target reduced power does not fall within the adjustable range of the power storage device 120.

[0045] (assignment) The following describes the problem. Here, we explain a case where target facilities participating in a DR request are selected based on the expected value of the reduction in power demand.

[0046] In such a case, the target facility may include a facility 100 having a large-scale power storage device 120 and a facility 100 having a small-scale power storage device 120. The target facility may also include a large-scale power storage device 120 and a small-scale power storage device 120.

[0047] Therefore, it is expected that the probability distribution of the predicted results of power reduction (hereinafter referred to as predicted power reduction) for the entire target facility will be multi-peaked, as shown in Figure 4. In Figure 4, the probability distribution of predicted power reduction is represented in a coordinate space defined by the probability distribution (vertical axis) and the predicted power reduction (horizontal axis).

[0048] Here, if target facilities are selected so that the expected value of the predicted power reduction approaches the target power reduction, the target power reduction falls within the success judgment range. However, because the probability distribution of the predicted power reduction is multi-modal, the expected value of the predicted power reduction may deviate from the most frequent value of the predicted power reduction.

[0049] If the most frequent value of the predicted power reduction is the actual power reduction, the target power reduction will not fall within the adjustable range, making it impossible to respond to DR requests and maintaining the power supply and demand balance in the power grid 12.

[0050] In the embodiment, in order to solve such a problem, a combination of target facilities is selected so that the combination of probability distributions of reduced power of demand power is unimodal.

[0051] (Selection of combination of target facilities) Selection of a combination of target facilities according to the embodiment will be described below. In Figures 5 and 6, the probability distribution of predicted power reduction is shown in a coordinate space defined by the probability distribution (vertical axis) and the predicted power reduction (horizontal axis).

[0052] For example, in a case where the probability distribution of power reductions for facilities #A to #C is predicted to be the probability distribution shown in Fig. 5, the target facilities are selected so that the combination of the probability distributions is unimodal. For example, facility #A and facility #B may be selected as the target facilities.

[0053] Here, a unimodal distribution does not only mean a distribution in which no peaks other than the mode exist, but also means a distribution in which peaks other than the mode are allowed if they satisfy specific conditions. The specific conditions may include a condition that the probability density of peaks other than the mode is equal to or less than a specific value (e.g., 20% of the probability density of the mode), or a condition that the predicted power reduction of peaks other than the mode is within a specific range (e.g., ±10% of the predicted power reduction of the mode).

[0054] Under these assumptions, it is preferable that the combination of target facilities is selected so that the most frequent value of the combination of probability distributions falls within a certain range (success determination range) based on the target power reduction, as shown in Fig. 6. It is preferable that the combination of target facilities is selected so that the target power reduction falls within the adjustable range of the power storage devices 120 installed in the target facilities.

[0055] The success determination range may be a certain range based on the target power reduction (for example, ±10% of the target power reduction), and the adjustable range may be a certain range based on the most frequent value of the reducible power (for example, a range with 90% reliability).

[0056] Although not particularly limited, a calculation example regarding a combination of target facilities will be described.

[0057] First, the probability distribution of the predicted power reduction of each facility 100 in a certain DR period t may be expressed by the following equation:

[0058]

number

[0059] Second, due to the reproducibility of the Gaussian distribution function, the probability distribution of the total predicted reduced power corresponding to a combination of target facilities in a certain DR period t may be expressed by the following equation:

[0060]

number

[0061] Here, for all of the local Gaussian distributions constituting the probability distribution of the total predicted reduced power, if the difference between the mean value characterizing the local Gaussian distribution and the target reduced power can be kept within the success determination range, the probability distribution of the total predicted reduced power can be made unimodal. The difference (loss function) between the mean value characterizing the local Gaussian distribution and the target reduced power may be expressed by the following formula.

[0062]

number

[0063]

number

[0064] When the priority of the facility 100 is taken into consideration, the difference (loss function) between the average value characterizing the local Gaussian distribution and the target reduced power consumption may be expressed by the following equation.

[0065]

number

[0066] The hyperparameter α may be determined empirically or by machine learning, such as AI, which may involve learning the difference between the target power reduction and the power reduction achieved by selecting a combination of target facilities according to the embodiment.

[0067] It should be noted that the above-described method for minimizing the loss function is merely an example, and other minimization methods may be employed. Specifically, a method for finding the lowest energy state of an Ising model may be used as the minimization method. For example, the Monte Carlo method may be used as the minimization method. A method for searching for an approximate solution to the problem of finding the ground state of a quantum Ising model with spin 1 / 2 by quantizing the loss function may be used. Examples of such methods include the exact diagonalization method, a method using a tensor network format, and quantum annealing technology.

[0068] (Power management method) A power management method according to an embodiment will be described below. Fig. 7 illustrates a case where the AC and the RA are separate entities. The above-mentioned power management server 200 may be the RA.

[0069] 7, in step S10, the RA receives information from the facility 100. The information may include the power demand of the facility 100, an operation plan for each distributed power source (the solar cell device 110, the power storage device 120, and the fuel cell device 130), an operation plan for the load device 140, and the like.

[0070] In step S11, the RA predicts the probability distribution of power for two or more facilities 100.

[0071] In step S12, the RA selects a combination of target facilities so that the combination of probability distributions is unimodal. The RA may select a combination of target facilities so that the most frequent value of the combination of probability distributions falls within the success determination range. The RA may select a combination of target facilities so that the target reduced power falls within the adjustable range of the power storage device 120.

[0072] In step S13, the RA evaluates whether to participate in the control message (DR request) received from the AC based on the combination of probability distributions corresponding to the combination of target facilities.

[0073] In step S14, the RA transmits to the AC whether or not it will participate in the DR request (participation availability). Here, the explanation will continue for the case of participating in the DR request.

[0074] In step S20, the AC sends a control message (DR request) to the RA.

[0075] In step S21, the RA transmits a control command to the target facility selected in step S12. The control command is a command for achieving the target power reduction. The control command may be a command for the EMS 160 or a command for the power storage device 120.

[0076] In step S22, the facility 100 controls the power storage device 120 in response to the control command.

[0077] In step S23, the RA receives from the facility 100 the control results (performance) in response to the control command.

[0078] (Action and effect) In the embodiment, the power management server 200 predicts a probability distribution of predicted reduced power and selects a combination of target facilities that contribute to the power supply and demand balance of the power grid 12 so that the combination of the probability distribution is unimodal. With this configuration, because the combination of the probability distribution is unimodal, it is possible to prevent the expected value of predicted reduced power from deviating from the most frequent value of predicted reduced power, and it is possible to appropriately reduce the power demand of two or more facilities 100 managed by the power management server 200. In other words, it is possible to appropriately select target facilities that contribute to reducing power demand.

[0079] [Other embodiments] Although the present invention has been described by the above-mentioned embodiments, the descriptions and drawings that form part of this disclosure should not be understood to limit the present invention. From this disclosure, various alternative embodiments, examples, and operating techniques will become apparent to those skilled in the art.

[0080] In the above disclosure, a case has been exemplified in which the power related to two or more facilities 100 is reduced power of the power demand of two or more facilities 100. However, the above disclosure is not limited to this. The power related to two or more facilities 100 may be increased power of the power demand of two or more facilities 100. Reduced power and increased power may be referred to as regulated power. The power related to two or more facilities 100 may be the power demand (flow power) of two or more facilities 100 or the output power (reverse flow power) of two or more facilities 100. The power related to two or more facilities 100 may be power generated by a distributed power source (e.g., solar cell device 110) installed in two or more facilities 100.

[0081] In the above disclosure, the balance of power supply and demand in the power grid 12 is adjusted by controlling the power storage device 120. However, the above disclosure is not limited to this. The balance of power supply and demand in the power grid 12 may be adjusted by a distributed power source other than the power storage device 120 (for example, the fuel cell device 130), or may be adjusted by controlling the load device 140.

[0082] Although not specifically mentioned in the above disclosure, the operation plans of the distributed power sources may be formulated autonomously by the EMS 160 or may be formulated centrally by the power management server 200. Similarly, the operation plans of the load devices 140 may be formulated autonomously by the EMS 160 or may be formulated centrally by the power management server 200.

[0083] Although not specifically mentioned in the above disclosure, the power storage device 120 that adjusts the balance of power supply and demand in the power grid 12 may be controlled autonomously by the EMS 160 or may be centrally controlled by the power management server 200.

[0084] Although not specifically mentioned in the above disclosure, a program may be provided that causes a computer to execute each process performed by the power management server 200. The program may also be recorded on a computer-readable medium. Using the computer-readable medium, the program can be installed on a computer. Here, the computer-readable medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, and may be, for example, a recording medium such as a CD-ROM or a DVD-ROM.

[0085] Alternatively, a chip may be provided that is configured by a memory that stores programs for executing the processes performed by the power management server 200 and a processor that executes the programs stored in the memory.

[0086] [Note] The above disclosure may be expressed as follows:

[0087] The first feature is that the power management device includes a management unit that manages two or more facilities connected to a power grid, and a control unit that selects a combination of target facilities that contributes to the power supply and demand balance of the power grid, wherein the control unit predicts a probability distribution of power for the two or more facilities and selects the combination of target facilities so that the combination of the probability distribution is unimodal.

[0088] A second feature is the power management device of the first feature, wherein the control unit selects the combination of the target facilities so that the most frequent value of the combination of the probability distribution falls within a certain range based on a target power.

[0089] A third feature is a power management device according to the first or second feature, wherein the control unit predicts a power range that can be adjusted by a distributed power source installed in the target facility, and selects a combination of the target facilities so that a target power falls within the power range.

[0090] A fourth feature is the power management device according to any one of the first to third features, wherein the power related to the two or more facilities is regulated power for adjusting power demands of the two or more facilities.

[0091] A fifth feature is a power management method including: a step A of managing two or more facilities connected to an electric power grid; and a step B of selecting a combination of target facilities that contributes to the balance of power supply and demand in the electric power grid, wherein the step B includes a step of predicting a probability distribution of power for the two or more facilities, and selecting the combination of the target facilities so that the combination of the probability distribution is unimodal. [Explanation of symbols]

[0092] 1...power management system, 11...network, 12...power system, 100...facility, 110...solar cell device, 120...power storage device, 130...fuel cell device, 140...load device, 160...EMS, 190...measuring device, 200...power management server, 210...communication unit, 220...management unit, 230...control unit

Claims

1. a management department that manages two or more facilities connected to the power grid; a control unit that selects a combination of target facilities that contributes to the power supply and demand balance of the power grid, the control unit predicts a probability distribution of power for the two or more facilities, and selects a combination of the target facilities such that the combination of the probability distributions is unimodal; A power management device, wherein the probability distribution of power for the two or more facilities includes a probability distribution that is not a normal distribution.

2. The power management device described in claim 1, wherein the two or more facilities are facilities having distributed power sources of different scales, or at least one of the two or more facilities has distributed power sources of different scales.

3. A power management device as described in claim 1, wherein the probability distribution of power for the two or more facilities is predicted by learning parameters including the past power demand of each of the two or more facilities.

4. The power management device according to claim 1 , wherein the control unit selects the combination of the target facilities so that the most frequent value of the combination of the probability distributions falls within a certain range based on a target power.

5. The power management device according to claim 1 , wherein the control unit predicts a power range that can be adjusted by distributed power sources installed in the target facilities, and selects a combination of the target facilities so that the target power falls within the power range.

6. The power management device according to claim 1 , wherein the power related to the two or more facilities is regulated power for adjusting the power demand of the two or more facilities.

7. Step A of managing two or more facilities connected to an electric power grid; and step B of selecting a combination of target facilities that contributes to the power supply and demand balance of the power system, Step B includes a step of predicting a probability distribution of power for the two or more facilities and selecting a combination of the target facilities such that the combination of the probability distributions is unimodal; A power management method, wherein the probability distribution of power for the two or more facilities includes a probability distribution that is not a normal distribution.

Images