Power management device and power management system

Abstract

To efficiently supply power to the load. [Solution] In the power management device, the power consumption forecasting unit creates a predicted consumption time period T200 that predicts the predicted power consumption of the load for each future elapsed time. The first calculation unit calculates the grid conversion efficiency R1 of the grid power source for each future elapsed time, and the second calculation unit calculates the specific conversion efficiency R2 of a specific power source for each future elapsed time. The identification unit identifies the time period within the predicted consumption time period T200 in which the specific conversion efficiency R2 is equal to or greater than the grid conversion efficiency R1 as the specific time period T300. The planning unit formulates a supply plan P100 so that power is supplied from the specific power source to the load for each future elapsed time in the specific time period T300.

Description

Technical Field

[0001] The present invention relates to a power management device and a power management system.

Background Art

[0002] For example, Japanese Patent Application Laid-Open No. 2021-90258 discloses a power system including a system power source and a specific power source provided separately from the system power source, and performing power supply to a load using the system power source and the specific power source. The power system includes a first variable unit that varies a first supply power supplied to the load using the system power source, a second variable unit that varies a second supply power supplied to the load using the specific power source, and a control unit that controls the first variable unit and the second variable unit so that power is supplied to the load from both the system power source and the specific power source.

[0003] The control unit derives a first set power and a second set power based on introduction parameters including a target supply power to the load, a first efficiency that is the efficiency of power supply of the system power source, and a second efficiency that is the efficiency of power supply of the specific power source. Further, the control unit controls the first variable unit so that a first supply power corresponding to the first set power is supplied to the load, and controls the second variable unit so that a second supply power corresponding to the second set power is supplied to the load. By this, it is said that power can be efficiently supplied to the load using the system power source and the specific power source.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Incidentally, the inventor of this application is considering supplying power to a load efficiently using both grid power and a specific power source, by a control system different from the power system disclosed in Patent Document 1 mentioned above. [Means for solving the problem]

[0006] The power management device disclosed herein comprises a recording unit, a power consumption forecasting unit, a first calculation unit, a second calculation unit, a specification unit, and a planning unit. The recording unit records the past power consumption of the load for each past elapsed time. The power consumption forecasting unit creates a predicted power consumption time period by forecasting the predicted power consumption of the load for each future elapsed time, based on the past power consumption for each past elapsed time recorded in the recording unit. The first calculation unit calculates the grid conversion efficiency, which is the power conversion efficiency for each future elapsed time, when the predicted power consumption is supplied to the load from the grid power supply for each future elapsed time during the predicted power consumption time period. The second calculation unit calculates the specific conversion efficiency, which is the power conversion efficiency for each future elapsed time, when the predicted power consumption is supplied to the load from a specific power supply, which is provided separately from the grid power supply, during the predicted power consumption time period. The specification unit specifies the time period within the predicted power consumption time period in which the specific conversion efficiency is equal to or greater than the grid conversion efficiency as a specific time period. The planning unit formulates a supply plan that indicates whether to supply power from the grid power source or the specific power source for each future elapsed time in the predicted consumption period, so that the predicted amount of power consumed is supplied to the load from the specific power source for each future elapsed time in the specified time period.

[0007] According to the power management device disclosed herein, during specific time periods within the predicted consumption period when the specific conversion efficiency is equal to or greater than the grid conversion efficiency, power can be supplied to the load efficiently by supplying power from a specific power source to the load. [Brief explanation of the drawing]

[0008] [Figure 1]Figure 1 is a conceptual diagram showing a power management system according to an embodiment. [Figure 2] Figure 2 is a block diagram of the power management system according to this embodiment. [Figure 3] Figure 3 is a flowchart illustrating the procedure for formulating a supply plan. [Figure 4] Figure 4 shows an example of past consumption time periods. [Figure 5] Figure 5 shows an example of a predicted consumption time period. [Figure 6] Figure 6 shows an example of a learning model generated by machine learning. [Figure 7] Figure 7 illustrates machine learning. [Figure 8] Figure 8 is a graph showing the difference between grid conversion efficiency and specific conversion efficiency depending on the amount of electricity supplied. [Modes for carrying out the invention]

[0009] Hereinafter, an embodiment of a power management system equipped with the power management device disclosed herein will be described with reference to the drawings. The embodiment described herein is, of course, not intended to particularly limit the present invention. Unless otherwise specified, the present invention is not limited to the embodiment described herein. Furthermore, components and parts that perform the same function will be appropriately denoted by the same reference numerals, and redundant descriptions will be omitted as appropriate.

[0010] Figure 1 is a conceptual diagram showing the power management system 100 according to this embodiment. The power management system 100 according to this embodiment is a system that manages the electricity consumed at the user's owned facility 5. The user referred to here is a customer of the management company that manages the power management system 100, and is a user registered with the power management system 100.

[0011] Owned facility 5 is a facility owned by the user. Here, owned facility 5 is a facility used by the user. For example, owned facility 5 is a residence owned by the user. The term "residence" here is not particularly limited to whether the user lives there or not; for example, it could be a residence (in other words, a building) for the user to live in, or it could be a rental property. However, owned facility 5 is not limited to a residence. Owned facility 5 could be, for example, an office or company building operated by the user.

[0012] In this embodiment, the power management system 100 is implemented, for example, by a client-server system. However, the power management system 100 may also be implemented by cloud computing. As shown in Figure 1, the power management system 100 includes a load 8, a grid power supply 10, a charge / discharge device 20, a renewable energy generation device 30, a power storage device 40, a control controller 50, and a power management device 70.

[0013] Load 8 is something that consumes electricity. Load 8 is, for example, something that uses electricity as its power source. The type of Load 8 is not particularly limited. Load 8 is, for example, household appliances such as televisions, refrigerators, and vacuum cleaners. Here, Load 8 is located in the owned facility 5. There may be multiple Load 8s. Power is supplied to Load 8, for example, through so-called outlets installed in the buildings within the owned facility 5.

[0014] The grid power supply 10 is a source of electricity. The grid power supply 10 is a source of electricity that supplies power from, for example, the commercial grid (for example, a power company). The electricity supplied from the grid power supply 10 is the electricity that users buy, and is known as purchased electricity.

[0015] The charge / discharge device 20 charges and discharges the electric vehicle 6. The electric vehicle 6 is, for example, a vehicle owned or used by a user. The electric vehicle 6 is equipped with a secondary battery 7. The secondary battery 7 is capable of repeated charging and discharging by the movement of charge carriers between a pair of electrodes (e.g., positive and negative electrodes) via an electrolyte, for example. As the secondary battery 7, for example, a lithium-ion secondary battery or a nickel-metal hydride battery may be used. In this embodiment, the secondary battery 7 is a lithium-ion secondary battery. Here, the electric vehicle 6 is a vehicle that uses the secondary battery 7 as a power source. The electric vehicle 6 is an electric vehicle that uses electricity as a power source, such as an electric vehicle, a hybrid vehicle, or a plug-in hybrid vehicle. The electric vehicle 6 may be a four-wheeled vehicle or a two-wheeled vehicle. Note that charging and discharging of the electric vehicle 6 here refers to charging and discharging the secondary battery 7 installed in the electric vehicle 6.

[0016] The charge / discharge device 20 is installed, for example, in the parking lot of the owned facility 5. The charge / discharge device 20 charges and discharges the electric vehicle 6 (more specifically, the secondary battery 7 mounted on the electric vehicle 6) parked in the parking lot. For example, the charge / discharge device 20 has a connection plug (not shown) that is connected to the electric vehicle 6. By connecting the connection plug to the electric vehicle 6, the charge / discharge device 20 becomes capable of charging and discharging the electric vehicle 6 connected to the connection plug. The number of charge / discharge devices 20 installed in the owned facility 5 is not particularly limited. In this embodiment, there is one charge / discharge device 20 installed in the owned facility 5, but there may be multiple.

[0017] The renewable energy power generation device 30 is a device that generates electric power using renewable energy. Here, the electric power generated by renewable energy is also referred to as renewable energy power. Examples of the energy sources of renewable energy include sunlight, wind power, hydraulic power, geothermal energy, solar heat, heat existing in the atmosphere or nature, biomass, and the like. The energy source of renewable energy by the renewable energy power generation device 30 is not particularly limited. In the present embodiment, the renewable energy power generation device 30 is a solar power generation device that uses sunlight as an energy source. Here, the renewable energy power generation device 30 has a solar panel (not shown) that receives sunlight. The renewable energy power generation device 30 is, for example, a device installed in the owned facility 5 or a device owned by a user.

[0018] In the present embodiment, the power supply source provided separately from the grid power source 10 is referred to as a specific power source 60. The power management system 100 includes the specific power source 60. The specific power source 60 is a supply source that supplies electric power obtained by a method different from that of the grid power source 10. The electric power supplied from the specific power source 60 is, for example, electric power different from the purchased power bought from an electric power company. In the present embodiment, the specific power source 60 includes an electric vehicle 6 connected to the charge-discharge device 20 and a renewable energy power generation device 30. Hereinafter, the discharge from the electric vehicle 6 means the discharge from the electric vehicle 6 connected to the charge-discharge device 20. The specific power source 60 has the electric vehicle 6 and the renewable energy power generation device 30. Here, the electric power supplied from the specific power source 60 can be the electric power discharged from the electric vehicle 6 through the charge-discharge device 20 or the renewable energy power (here, solar power) generated from the renewable energy power generation device 30. Note that the number of specific power sources 60 included in the power management system 100 is not particularly limited. Here, the number of specific power sources 60 is two, namely the electric vehicle 6 and the renewable energy power generation device 30, but may be three or more. Also, the number of specific power sources 60 may be only one of either the electric vehicle 6 or the renewable energy power generation device 30. That is, either the electric vehicle 6 or the renewable energy power generation device 30 may be omitted.

[0019] The power storage device 40 is a device that stores electric power. The power storage device 40 is connected to, for example, the charge / discharge device 20 and the renewable energy power generation device 30. The electric power discharged from the electric vehicle 6 to the charge / discharge device 20 is stored in the power storage device 40. Also, the renewable energy power generated by the renewable energy power generation device 30 is stored in the power storage device 40. Incidentally, the power storage device 40 may be connected to the utility power source 10. The electric power supplied from the utility power source 10 may be stored in the power storage device 40. Incidentally, the number of power storage devices 40 may be one or plural. For example, when there are plural power storage devices 40, the electric power discharged from the electric vehicle 6, the electric power generated by the renewable energy power generation device 30, and the electric power supplied from the utility power source 10 may be stored in separate power storage devices 40 respectively.

[0020] The owner controller 50 controls the supply of electric power within the owned facility 5. The owner controller 50 controls the supply of the electric power from the utility power source 10, the electric power discharged from the electric vehicle 6, and the electric power generated by the renewable energy power generation device 30 to the load 8. The owner controller 50 controls the supply of the electric power stored in the power storage device 40 to the load 8. When supplying electric power to the load 8, the owner controller 50 controls the selection of the power supply source (for example, any one of the utility power source 10, the electric vehicle 6 connected to the charge / discharge device 20, and the renewable energy power generation device 30), the amount of electric power supplied to the load 8, the timing of supplying electric power to the load 8, etc. The owner controller 50 is a general term for a so-called smart meter (an electronic watt-hour meter having a function of digitally measuring electric power and a communication function), a power conditioner for the charge / discharge device 20, a power conditioner for the renewable energy power generation device 30, and a power conditioner for the power storage device 40, and is a controller that integrates these.

[0021] As shown in Figure 1, the owner controller 50 is communicatively connected to, for example, the grid power supply 10, the charge / discharge device 20, the renewable energy generator 30, and the energy storage device 40. The configuration of the owner controller 50 is not particularly limited. The owner controller 50 is, for example, a microcomputer. The owner controller 50 includes, for example, an interface, a CPU, ROM, and RAM. The owner controller 50 may consist of a single computer or multiple computers. Furthermore, the owner controller 50 may be implemented by, for example, a personal computer.

[0022] The power management device 70 is a device that manages the power supplied to the owned facility 5 (e.g., load 8). For example, if the power management system 100 is implemented by cloud computing, the power management device 70 functions as a so-called cloud server. For example, if the power management system 100 is implemented by a client-server system, the power management device 70 functions as a server. The power management device 70 includes, for example, an interface, a CPU, ROM, and RAM. The power management device 70 may consist of a single computer (e.g., a single server) or multiple computers (e.g., multiple servers).

[0023] In this embodiment, as shown in Figure 1, the power management device 70 is connected to the owning controller 50 of the owning facility 5 in a communicative manner. Here, the power management device 70 is connected to the owning controller 50 via the Internet. For example, the power management device 70 is configured or programmed to send and receive information to and from the owning controller 50. The power management device 70 may also be connected to the electric vehicle 6, the grid power supply 10, the charge / discharge device 20, the renewable energy generation device 30, and the energy storage device 40 in a communicative manner.

[0024] The configuration of the power management system 100 according to this embodiment has been described above. Incidentally, the electricity supplied from the grid power source 10 is, for example, purchased electricity from a power company. On the other hand, the renewable energy electricity generated by the renewable energy power generation device 30 is cheaper than the electricity supplied from the grid power source 10. Also, the electricity charged to the electric vehicle 6 is electricity that has already been purchased, so there will be no cost even if it is consumed in the future. For this reason, in a facility 5 that owns specific power sources 60 such as the renewable energy power generation device 30 and the electric vehicle 6, as in this embodiment, it is preferable that the electricity consumed by the load 8 comes from the specific power sources 60 as much as possible. Therefore, in this embodiment, the power management device 70 manages the power so that the load 8 consumes electricity from the specific power sources 60 rather than from the grid power source 10.

[0025] Figure 2 is a block diagram of the power management system 100 according to this embodiment. In this embodiment, as shown in Figure 2, the power management device 70 includes a storage unit 71, a recording unit 81, a power consumption prediction unit 83, a first calculation unit 85, a second calculation unit 87, a specific unit 91, a planning unit 93, and a plan transmission unit 95. Each part of the power management device 70 may be implemented by software or by hardware. Furthermore, each part of the power management device 70 may be implemented by one or more processors or by circuits.

[0026] In this embodiment, a supply plan P100 (see Figure 3) is formulated that indicates which power source, grid power 10 or specific power source 60, will be supplied to the load 8 during any given time period. The procedure for formulating the supply plan P100 will be explained below in accordance with the flowchart in Figure 3.

[0027] Figure 4 shows an example of a past consumption time period T100. In this embodiment, prior to formulating the supply plan P100, the recording unit 81 in Figure 2 records the past consumption time period T100 (see Figure 4). Here, the past consumption time period T100 is data that records the past power consumption amount V10 of the load 8 for each past elapsed time T10. Past elapsed time T10 is the time between any two points in time. The specific numerical value of past elapsed time T10 is not particularly limited, but for example, past elapsed time T10 is 1 hour. Past power consumption amount V10 is the amount of power that the load 8 actually consumed during the past elapsed time T10 (for example, 1 hour). The recording unit 81 records the past power consumption amount V10 for each past elapsed time T10.

[0028] The length of the past consumption time period T100 (referred to here as the period) is not particularly limited. As will be explained in more detail later, the predicted consumption time period T200 (see Figure 5) is predicted based on the past consumption time period T100. For example, if the period of the past consumption time period T100 is long, the predicted consumption time period T200 can be predicted with high accuracy, but it may require processing time. Therefore, the period of the past consumption time period T100 should be set considering the accuracy of the predicted consumption time period T200 and the processing time required to predict the predicted consumption time period T200. In addition, the past consumption time period T100 may include weather information for each past elapsed time period T10. Weather information may include, for example, weather conditions (sunny, rainy, cloudy, etc.), temperature, precipitation, etc. for each past elapsed time period T10.

[0029] For example, the owned controller 50 (see Figure 1) in owned facility 5 records and stores past consumption time periods T100 (past power consumption V10 of load 8 for each past elapsed time T10). Therefore, the owned controller 50 transmits the stored past consumption time periods T100 to the power management device 70. The recording unit 81 acquires the past consumption time periods T100 transmitted from the owned controller 50. The past power consumption V10 for each past elapsed time T10 recorded by the recording unit 81, i.e., the past consumption time periods T100, are stored in the storage unit 71.

[0030] In this embodiment, with the past consumption time period T100 recorded by the recording unit 81, the flowchart in Figure 3 is executed sequentially. Here, first, in step S101 in Figure 3, the power consumption prediction unit 83 in Figure 2 predicts the predicted consumption time period T200. Figure 5 is a diagram showing an example of the predicted consumption time period T200. As shown in Figure 5, the predicted consumption time period T200 here refers to the predicted power consumption amount V20 of the load 8 for each future elapsed time T20. Future elapsed time T20 refers to the elapsed time in the future. The specific value of future elapsed time T20 is the same value as the past elapsed time T10 above, for example, every hour. The predicted power consumption amount V20 is the amount of power that the load 8 is predicted to consume in the future elapsed time T20 (for example, 1 hour).

[0031] The method by which the power consumption prediction unit 83 predicts the predicted consumption time period T200 is not particularly limited. In this embodiment, the power consumption prediction unit 83 predicts the predicted consumption time period T200 based on the past power consumption amount V10 for each past elapsed time T10 recorded by the recording unit 81 (i.e., past consumption time period T100). The power consumption prediction unit 83 predicts the predicted consumption time period T200 using machine learning. In this embodiment, as shown in Figure 2, the power consumption prediction unit 83 has a model generation unit 83a and a machine learning unit 83b.

[0032] Figure 6 shows an example of a learning model MD1 generated by machine learning. Here, the model generation unit 83a generates a learning model MD1 as shown in Figure 6. Here, the model generation unit 83a generates the learning model MD1 using past consumption time periods T100 as training data. For example, the power consumption of load 8 may be high while a user is staying at their facility 5. Also, depending on the user, there may be times of day when the power consumption of load 8 is high and times of day when it is low. For example, depending on the user, the power consumption of load 8 may be higher in the morning and evening than in the daytime or midnight. In other words, there may be a correlation between the date and time t11 for each past elapsed time T10 and the past power consumption V10. For example, the model generation unit 83a takes the date and time t11 for each past elapsed time T10 from the past consumption time periods T100 as input and outputs the past power consumption V10 for each date and time T10.

[0033] However, as shown in Figure 6, the data input to the learning model MD1 may be the date and time t11 for each past elapsed time T10 and past weather information W10. For example, on days with low temperatures or high temperatures, the power consumption of load 8 (e.g., air conditioner) may be high. Therefore, there may be a correlation between past weather information W10 and past power consumption V10. In this embodiment, the model generation unit 83a may generate the learning model MD1 by taking the date and time t11 for each past elapsed time T10 and past weather information W10 for each past elapsed time T10 as input and past power consumption V10 for each past elapsed time T10 as output. However, the past weather information W10 may be omitted when generating the learning model MD1.

[0034] Figure 7 is a diagram illustrating machine learning. Next, as shown in Figure 7, the machine learning unit 83b outputs the predicted power consumption V20 at date and time t21 for each future elapsed time T20. The machine learning unit 83b uses the learning model MD1 to input the date and time t21 for each future elapsed time T20 into the learning model MD1 and outputs the predicted power consumption V20. The predicted power consumption V20 output from the learning model MD1 becomes the predicted power consumption for each future elapsed time T20. Note that when past weather information W10 is used as input data for the learning model MD1 generated by the model generation unit 83a, as shown in Figure 6, the machine learning unit 83b may input the date and time t21 for each future elapsed time T20 and the future weather information W20 into the learning model MD1 and output the predicted power consumption V20, as shown in Figure 7. Note that future weather information W20 can be obtained, for example, from a weather information provider. Here, the power consumption forecasting unit 83 obtains future weather information W20 for each future elapsed time T20 from a server operated by a weather information provider. Note that if past weather information W10 is omitted in Figure 6, then future weather information W20 will be omitted in Figure 7.

[0035] In this embodiment, the model generation unit 83a may add the predicted power consumption V20 for each future elapsed time T20 output by the machine learning unit 83b to the training data to generate a new learning model MD1. Alternatively, if the past power consumption time period T100 is updated and there is a newly added past power consumption V10 for each past elapsed time T10 of load 8 in the past power consumption time period T100, the model generation unit 83a may add the newly added past power consumption V10 for each past elapsed time T10 of load 8 to the training data to generate a new learning model MD1.

[0036] In this embodiment, the power consumption prediction unit 83 can create a predicted power consumption time period T200 using the predicted power consumption V20 for each future elapsed time T20 obtained from machine learning. The predicted power consumption time period T200 predicted by the power consumption prediction unit 83 is stored in the storage unit 71.

[0037] Next, in step S103 of Figure 3, the first calculation unit 85 of Figure 2 calculates the grid conversion efficiency R1 in the grid power source 10. Grid conversion efficiency R1 is the efficiency of converting the power output from the grid power source 10 into power for supply to the load 8. Grid conversion efficiency R1 is, for example, the ratio of the amount of power actually supplied to the load 8 to the amount of power output from the grid power source 10. When grid conversion efficiency R1 is high, the power conversion loss is small, and power can be supplied to the load 8 efficiently. On the other hand, when grid conversion efficiency R1 is low, the power conversion loss is large, and power cannot be supplied to the load 8 efficiently. Grid conversion efficiency R1 is calculated according to the grid supply power amount, which is the amount of power supplied from the grid power source 10 at one time. Grid supply power amount is the amount of power supplied from the grid power source 10 per unit time. For example, when the grid supply power amount is large, the grid conversion efficiency R1 tends to be high, and when the grid supply power amount is small, the grid conversion efficiency R1 tends to be low. In this embodiment, the storage unit 71 has a system efficiency conversion table TB10 (see Figure 2) pre-stored, which shows the system conversion efficiency R1 according to the amount of power supplied to the grid. The first calculation unit 85 calculates the system conversion efficiency R1 by applying the amount of power supplied to the grid to the system efficiency conversion table TB10. Here, the first calculation unit 85 calculates the system conversion efficiency R1 for each amount of power supplied to the grid.

[0038] In this embodiment, the first calculation unit 85 calculates the grid conversion efficiency R1 for each future elapsed time T20 in the predicted consumption time period T200. Here, the first calculation unit 85 calculates the grid conversion efficiency R1 for each future elapsed time T20, assuming that the same amount of grid-supplied power as the predicted power consumption V20 is supplied from the grid power source 10 to the load 8. The first calculation unit 85 calculates the grid conversion efficiency R1 such that the grid conversion efficiency R1 increases as the predicted power consumption V20 increases. The grid conversion efficiency R1 for each future elapsed time T20 calculated by the first calculation unit 85 is stored in the storage unit 71.

[0039] Next, in step S105 of Figure 3, the second calculation unit 87 of Figure 2 calculates the specific conversion efficiency R2 for the specific power source 60 (here, the specific power source 60 including the renewable energy power generation device 30 and the electric vehicle 6). The specific conversion efficiency R2 is the efficiency of converting the power output from the specific power source 60 into power for supply to the load 8. The specific conversion efficiency R2 is, for example, the ratio of the amount of power actually supplied to the load 8 to the amount of power output from the specific power source 60. In the case of multiple specific power sources 60 as in this embodiment, the average of the conversion efficiencies of each specific power source 60 may be the specific conversion efficiency R2. Here, when the specific conversion efficiency R2 is high, it can be said that the power conversion loss is small and power can be efficiently supplied from the specific power source 60 to the load 8. On the other hand, when the specific conversion efficiency R2 is low, it can be said that the power conversion loss is large and power cannot be efficiently supplied from the specific power source 60 to the load 8. The specific conversion efficiency R2 is calculated according to the specific power supply amount, which is the amount of power supplied at one time output from the specific power source 60. The specified power supply amount is the amount of power supplied from a specific power source 60 per unit time. Similar to the grid conversion efficiency R1, for example, when the specified power supply amount is large, the specified conversion efficiency R2 tends to be high, and when the specified power supply amount is small, the specified conversion efficiency R2 tends to be low. In this embodiment, the storage unit 71 has a specified efficiency conversion table TB20 (see Figure 2) that shows the specified conversion efficiency R2 according to the specified power supply amount stored in advance. The second calculation unit 87 calculates the specified conversion efficiency R2 by applying the specified power supply amount to the specified efficiency conversion table TB20. Here, the second calculation unit 87 calculates the specified conversion efficiency R2 for each specified power supply amount.

[0040] In this embodiment, the second calculation unit 87 calculates a specific conversion efficiency R2 for each future elapsed time T20 in the predicted consumption time period T200. Here, the second calculation unit 87 calculates the specific conversion efficiency R2 for each future elapsed time T20, assuming that the same specific supply amount as the predicted power consumption V20 is supplied from the specific power source 60 to the load 8. The second calculation unit 87 calculates the specific conversion efficiency R2 such that the higher the predicted power consumption V20, the higher the specific conversion efficiency R2 becomes. The specific conversion efficiency R2 for each future elapsed time T20 calculated by the second calculation unit 87 is stored in the storage unit 71.

[0041] In this embodiment, the grid efficiency conversion table TB10 shown in Figure 2 is different from the specific efficiency conversion table TB20. That is, even with the same amount of power supplied, the grid conversion efficiency R1 and the specific conversion efficiency R2 may differ. Figure 8 is a graph showing the difference between the grid conversion efficiency R1 and the specific conversion efficiency R2 according to the amount of power supplied. Here, as shown in Figure 8, when the amount of power supplied is small (for example, less than the reference power amount NV1), the grid conversion efficiency R1 is higher than the specific conversion efficiency R2. On the other hand, when the amount of power supplied is large (for example, greater than or equal to the reference power amount NV1), the specific conversion efficiency R2 is about the same as or higher than the grid conversion efficiency R1. Therefore, for example, when the amount of power supplied to load 8 (predicted power consumption V20) is less than the reference power amount NV1, it is considered that supplying power to load 8 from the grid power source 10 is more efficient than supplying power from the specific power source 60. On the other hand, when the amount of electricity to be supplied to load 8 (predicted power consumption V20) is equal to or greater than the reference energy amount NV1, it is considered that supplying power to load 8 from the specific power source 60 is more efficient than supplying power from the grid power source 10. Here, the reference energy amount NV1 is the amount of electricity that marks the boundary between the grid conversion efficiency R1 and the specific conversion efficiency R2, which is higher.

[0042] Next, in step S107 of Figure 3, the identification unit 91 of Figure 2 identifies a specific time period T300 from the predicted consumption time period T200. This specific time period T300 is the time period during which power output from the specific power source 60 is supplied to the load 8. In other words, during the grid time period T400 (see Figure 5), which excludes the specific time period T300 from the predicted consumption time period T200, power output from the grid power source 10 will be supplied to the load 8. Note that the method for identifying the specific time period T300 is not particularly limited.

[0043] In this embodiment, the identification unit 91 identifies a specific time period T300 from the predicted consumption time period T200 based on the grid conversion efficiency R1 and specific conversion efficiency R2 for each future elapsed time T20. The identification unit 91 compares the grid conversion efficiency R1 and specific conversion efficiency R2 with respect to the predicted power consumption V20 for each future elapsed time T20. The identification unit 91 then identifies the time period corresponding to the future elapsed time T20 in the predicted consumption time period T200 where the specific conversion efficiency R2 is equal to or greater than the grid conversion efficiency R1 as the specific time period T300. Time periods corresponding to the future elapsed time T20 in which the specific conversion efficiency R2 is less than the grid conversion efficiency R1 are not included in the specific time period T300. Here, the identification unit 91 identifies the time period corresponding to the future elapsed time T20 in the predicted consumption time period T200 in which the specific conversion efficiency R2 is less than the grid conversion efficiency R1 as the grid time period T400.

[0044] As described above, when the specific conversion efficiency R2 is equal to or greater than the grid conversion efficiency R1, the amount of power supplied to the load 8 will be equal to or greater than the reference power amount NV1. Therefore, as shown in Figure 5, the identification unit 91 identifies the time period within the predicted consumption time period T200 in which the predicted power consumption amount V20 is equal to or greater than the reference power amount NV1 as the specific time period T300. Also, as described above, when the specific conversion efficiency R2 is less than the grid conversion efficiency R1, the amount of power supplied to the load 8 will be less than the reference power amount NV1. Therefore, as shown in Figure 5, the identification unit 91 identifies the time period within the predicted consumption time period T200 in which the predicted power consumption amount V20 is less than the reference power amount NV1 as the grid time period T400. The information regarding the specific time period T300 and grid time period T400 identified by the identification unit 91 is stored in the storage unit 71.

[0045] Next, in step S109 of Figure 3, the planning unit 93 of Figure 2 formulates the supply plan P100. The supply plan P100 is a plan of whether to supply power to the load 8 from the grid power source 10 or the specific power source 60. The supply plan P100 is a plan that shows whether to supply power from the grid power source 10 or the specific power source 60 at each future elapsed time T20 in the predicted consumption time period T200, so that in the specific time period T300, power equal to the predicted power consumption V20 will be supplied from the specific power source 60 at each future elapsed time T20. In other words, the supply plan P100 is a plan to control the specific power source 60 so that by the date and time of each future elapsed time T20 in the specific time period T300, the amount of energy stored in the specific power source 60 will be equal to or greater than the predicted power consumption V20 corresponding to that date and time. The amount of energy stored in the specific power source 60 referred to here is, for example, the amount of energy stored in the energy storage device 40, which stores electricity generated from the renewable energy power generation device 30 and electricity discharged from the electric vehicle 6.

[0046] In this embodiment, the amount of electricity that can be generated from the renewable energy power generation device 30 and the amount of electricity that can be discharged from the electric vehicle 6 may differ at each future elapsed time T20. Therefore, the planning unit 93 acquires predicted vehicle SOC data DT10 (see Figure 2) and predicted renewable energy electricity amount data DT20 (see Figure 2). Predicted vehicle SOC data DT10 refers to the State of Charge (SOC) of the secondary battery 7 installed in the electric vehicle 6 at each future elapsed time T20. SOC is an index that shows the battery capacity when the fully charged state of the secondary battery 7 is 100% and the completely discharged state is 0%. The method for predicting the predicted vehicle SOC data DT10 is not particularly limited and can be predicted by conventionally known methods. For example, the planning unit 93 predicts the predicted vehicle SOC data DT10 by machine learning based on past driving data of the electric vehicle 6. The above driving data is, for example, data that associates the distance traveled and SOC at each past elapsed time T10. Here, using driving data as training data, for example, the distance traveled every T10 in the past is taken as input, and the State of Operation (SOC) of the electric vehicle 6 every T10 in the past is taken as output to generate an SOC learning model. Then, the distance traveled every T20 in the future is input to the SOC learning model, and the SOC of the electric vehicle 6 every T20 in the future is output. The distance traveled every T20 in the future may be estimated from past driving data or from destination information entered by the user. In this way, the predicted vehicle SOC data DT10 is predicted. The predicted vehicle SOC data DT10 is stored in the storage unit 71, as shown in Figure 2. Therefore, the planning unit 93 can obtain the predicted vehicle SOC data DT10 from the storage unit 71.

[0047] The predicted renewable energy power amount data DT20 refers to the amount of renewable energy power generated from the renewable energy power generation device 30 at each future elapsed time T20. The method for predicting the predicted renewable energy power amount data DT20 is not particularly limited and can be predicted by conventionally known methods. For example, the planning unit 93 predicts the predicted renewable energy power amount data DT20 by machine learning based on past renewable energy power amount data. Here, past renewable energy power amount data is data that associates the amount of electricity generated from the renewable energy power generation device 30 at each past elapsed time T10 with the weather information at that time. Past renewable energy power amount data is stored, for example, in the owner controller 50. Therefore, the planning unit 93 can obtain past renewable energy power amount data from the owner controller 50. Here, using past renewable energy power amount data as training data, for example, past weather information W10 for each past elapsed time T10 is taken as input, and a renewable energy learning model is generated that outputs the amount of renewable energy power for each past elapsed time T10. Then, future weather information W20 for each future elapsed time T20 is input to the renewable energy learning model, and the amount of renewable energy for each future elapsed time T20 is output. In this way, the predicted renewable energy amount data DT20 is predicted. The predicted renewable energy amount data DT20 is stored in the storage unit 71, as shown in Figure 2. Therefore, the planning unit 93 can obtain the predicted renewable energy amount data DT20 from the storage unit 71.

[0048] In this embodiment, the planning unit 93 formulates a supply plan P100 based on the predicted vehicle SOC data DT10 and the predicted renewable energy power data DT20 so that by the date and time of each future elapsed time T20 in a specific time period T300, the amount of energy stored in the energy storage device 40 will be equal to or greater than the predicted power consumption V20 corresponding to that date and time. Here, for example, if it is possible to secure power equivalent to the predicted power consumption V20 from renewable energy (for example, if the predicted renewable energy power amount ≥ predicted power consumption V20), the supply plan P100 is formulated to supply power equivalent to the predicted power consumption V20 generated from the renewable energy power generation device 30 to the energy storage device 40 by the corresponding date and time. For example, if renewable energy cannot secure the predicted power consumption V20 (for example, if predicted renewable energy amount < predicted power consumption V20), it is determined whether the remaining predicted power consumption V20 can be secured by the power discharged from the electric vehicle 6 (for example, whether predicted renewable energy amount + predicted vehicle discharge amount ≥ predicted power consumption V20). If it can be secured (for example, if predicted renewable energy amount + predicted vehicle discharge amount ≥ predicted power consumption V20), a supply plan P100 is formulated to supply power equivalent to the predicted power consumption V20 to the energy storage device 40 by the corresponding date and time using the power generated from the renewable energy generator 30 and the power discharged from the electric vehicle 6. If it is estimated that the electricity generated by the renewable energy generator 30 and the electricity discharged from the electric vehicle 6 will not be sufficient to secure the predicted power consumption V20 by the corresponding date and time (when predicted renewable energy amount + predicted vehicle discharge amount < predicted power consumption V20), then the supply plan P100 should be designed so that the remaining predicted power consumption V20 is secured by electricity supplied from the grid power source 10. Note that the electricity supplied from the grid power source 10 has a predetermined price for each elapsed time. Therefore, when using electricity from the grid power source 10, the supply plan P100 should be optimized to use it during a relatively inexpensive time period. In the supply plan P100, electricity is supplied from the grid power source 10 to the load 8 during grid time period T400.

[0049] Furthermore, when formulating the supply plan P100, it is possible to use prediction algorithms such as autoregressive integrated moving average models and recurrent neural networks, or optimization algorithms such as mixed-integer linear programming (MILP) algorithms.

[0050] After the supply plan P100 is formulated by the planning unit 93 in this manner, the process proceeds to step S111 in Figure 3. In step S111, the plan transmission unit 95 in Figure 2 transmits the supply plan P100 to the owner controller 50.

[0051] The owner controller 50 receives the supply plan P100. In this embodiment, as shown in Figure 2, the owner controller 50 includes a supply control unit 51. After the owner controller 50 receives the supply plan P100, the supply control unit 51 of the owner controller 50 controls the energy storage device 40 to supply the stored power to the load 8 at the timing when the load 8 consumes power from the specific power source 60 (specific time period T300), based on the supply plan P100. The supply control unit 51 also controls the charge / discharge device to discharge from the electric vehicle 6 at the timing when the electric vehicle 6 discharges, based on the supply plan P100. Furthermore, the supply control unit 51 controls the grid power source 10 to supply power to the load 8 at the timing when power is supplied from the grid power source 10 (for example, the grid time period T400 within the predicted consumption time period T200), based on the supply plan P100.

[0052] As described above, in this embodiment, as shown in Figure 1, the power management system 100 comprises a grid power supply 10 that supplies power, a specific power supply 60 provided separately from the grid power supply 10, and a power management device 70. As shown in Figure 2, the power management device 70 comprises a recording unit 81, a power consumption prediction unit 83, a first calculation unit 85, a second calculation unit 87, a specific unit 91, and a planning unit 93. As shown in Figure 4, the recording unit 81 records the past power consumption V10 of the load 8 for each past elapsed time T10. Based on the past power consumption V10 for each past elapsed time T10 recorded in the recording unit 81, the power consumption prediction unit 83 creates a predicted power consumption time period T200 (see Figure 5) which predicts the predicted power consumption V20 of the load 8 for each future elapsed time T20. The first calculation unit 85 calculates the grid conversion efficiency R1 (see step S103 in Figure 3), which is the power conversion efficiency for each future elapsed time T20 when the predicted power consumption amount V20 is supplied from the grid power source 10 to the load 8 at each future elapsed time T20 during the predicted power consumption time period T200. The second calculation unit 87 calculates the specific conversion efficiency R2 (see step S105 in Figure 3), which is the power conversion efficiency for each future elapsed time T20 when the predicted power consumption amount V20 is supplied from the specific power source 60 to the load 8 at each future elapsed time T20 during the predicted power consumption time period T200. The identification unit 91 identifies the time period within the predicted power consumption time period T200 in which the specific conversion efficiency R2 is equal to or greater than the grid conversion efficiency R1 as the specific time period T300, as shown in step S107 in Figure 3. As shown in step S109 of Figure 3, the planning unit 93 formulates a supply plan P100 that indicates whether power will be supplied from the grid power 10 or the specific power 60 at each future elapsed time T20 in the predicted consumption time T200, so that in the specific time period T300, power of the predicted consumption amount V20 will be supplied from the specific power source 60 to the load 8 at each future elapsed time T20.

[0053] In this embodiment, during time periods when the specific conversion efficiency R2 is equal to or greater than the grid conversion efficiency R1, supplying power to the load 8 from the specific power source 60 is more efficient than supplying power from the grid power source 10. On the other hand, during time periods when the specific conversion efficiency R2 is less than the grid conversion efficiency R1, supplying power to the load 8 from the grid power source 10 is more efficient than supplying power from the specific power source 60. Therefore, during the specific time period T300 within the predicted consumption time period T200, when the specific conversion efficiency R2 is equal to or greater than the grid conversion efficiency R1, power can be efficiently supplied to the load 8 by supplying power from the specific power source 60 to the load 8. Power can be efficiently supplied to the load 8 by controlling the system 50 to supply power to the load 8 in accordance with the supply plan P100 formulated by the planning unit 93.

[0054] In this embodiment, the first calculation unit 85 calculates the grid conversion efficiency R1 such that the grid conversion efficiency R1 increases as the predicted power consumption V20 increases. The second calculation unit 87 calculates the specific conversion efficiency R2 such that the specific conversion efficiency R2 increases as the predicted power consumption V20 increases. Thus, the grid conversion efficiency R1 and the specific conversion efficiency R2 can decrease as the amount of power output at one time (here, the predicted power consumption V20) decreases. Therefore, the grid conversion efficiency R1 and the specific conversion efficiency R2 can be calculated according to the amount of power output at one time.

[0055] In this embodiment, as shown in Figure 8, when the predicted power consumption V20 (e.g., supplied power) is less than a predetermined reference power amount NV1, the grid conversion efficiency R1 is higher than the specific conversion efficiency R2. When the predicted power consumption V20 is equal to or greater than the reference power amount NV1, the specific conversion efficiency R2 is equal to or greater than the grid conversion efficiency R1. As shown in Figure 5, the identification unit 91 identifies the time period T300 within the predicted consumption time period T200 in which the predicted power consumption V20 is equal to or greater than the reference power amount NV1. In this way, the specific time period T300 in which power is supplied from the specific power source 60 to the load 8 can be identified based on the reference power amount NV1.

[0056] In this embodiment, as shown in Figure 2, the power consumption prediction unit 83 includes a model generation unit 83a and a machine learning unit 83b. As shown in Figure 6, the model generation unit 83a uses past power consumption V10 for each past elapsed time T10 as training data, takes the date and time t11 for each past elapsed time T10 as input, and generates a learning model MD1 with past power consumption V10 for each past elapsed time T10 as output. In the predicted consumption time period T200, as shown in Figure 7, the machine learning unit 83b inputs the date and time t21 for each future elapsed time T20 into the learning model MD1 and outputs a predicted power consumption V20 for each future elapsed time T20. For example, a user often has the same daily (e.g., weekday) cycle, and the time they spend at the owned facility 5 may be the same. The amount of power consumed by the load 8 may increase during the time the user is staying. Therefore, by using machine learning based on past power consumption V10 for each elapsed time T20, it becomes easier to predict the predicted power consumption V20 for each future elapsed time T20 in the predicted power consumption period T200.

[0057] In this embodiment, the model generation unit 83a adds the predicted power consumption V20 for each future elapsed time T20 output by the machine learning unit 83b to the training data to generate the learning model MD1. This improves the accuracy of machine learning, making it easier to accurately predict the predicted power consumption V20 for each future elapsed time T20 in the predicted power consumption time period T200.

[0058] As described above, this specification includes the disclosures set forth in the following sections.

[0059] Section 1: A recording unit that records the past power consumption of the load for each past elapsed time, A power consumption prediction unit creates predicted power consumption time periods by predicting the predicted power consumption of the load for each future elapsed time, based on the past power consumption for each past elapsed time recorded in the recording unit, A first calculation unit calculates the grid conversion efficiency, which is the power conversion efficiency for each future elapsed time, when the predicted amount of power consumption is supplied from the grid power source to the load at each future elapsed time during the predicted consumption time period. A second calculation unit calculates a specific conversion efficiency, which is the power conversion efficiency for each future elapsed time, when the predicted amount of power consumption is supplied to the load from a specific power source, which is provided separately from the grid power source, at each future elapsed time during the predicted consumption time period. A special unit identifies the time period within the predicted consumption time period in which the specified conversion efficiency is equal to or greater than the system conversion efficiency as a specific time period, A planning unit that, in the specified time period, devises a supply plan indicating whether to supply power from the grid power source or the specified power source for each future elapsed time period in the predicted consumption time period, so that the predicted amount of power consumption is supplied to the load from the specified power source at each future elapsed time period in the predicted consumption time period, A power management device equipped with this device.

[0060] Section 2: The first calculation unit calculates the grid conversion efficiency such that the grid conversion efficiency increases as the predicted power consumption increases. The power management device described in item 1, wherein the second calculation unit calculates the specific conversion efficiency such that the specific conversion efficiency increases as the predicted power consumption increases.

[0061] Section 3: When the predicted power consumption is less than a predetermined standard power consumption, the grid conversion efficiency becomes higher than the specified conversion efficiency. When the predicted power consumption is equal to or greater than the reference power consumption, the specific conversion efficiency becomes equal to or greater than the grid conversion efficiency. The power management device described in item 1 or 2, wherein the identifying unit identifies the time period within the predicted consumption time period in which the predicted power consumption amount is equal to or greater than the standard power amount as the identified time period.

[0062] Section 4: The aforementioned power consumption forecasting unit is A model generation unit generates a learning model that uses the past power consumption amounts for each past elapsed time as training data, takes the date and time for each past elapsed time as input, and outputs the past power consumption amounts for each past elapsed time. A machine learning unit inputs the date and time for each future elapsed time into the learning model during the predicted consumption time period and outputs the predicted power consumption for each future elapsed time. A power management device having any one of the features described in paragraphs 1 to 3.

[0063] Section 5: The power management device described in item 4, wherein the model generation unit adds the predicted power consumption amounts for each future elapsed time output by the machine learning unit to the training data to generate the learning model.

[0064] Item 6: A power management device described in any one of items 1 through 5, The aforementioned power supply system, The aforementioned specific power source, A power management system equipped with [specific features / equipment]. [Explanation of Symbols]

[0065] 6 Electric Vehicles 7 Secondary battery 10 Grid power supply 60 Specified power supply 70 Power management device 81 Records Section 83 Power Consumption Forecasting Unit 83a Model generation unit 83b Machine Learning Department 85 First Calculation Unit 87 Second Calculation Unit 91 Specific part 93 Planning Department 100 Power Management Systems R1 System Conversion Efficiency R2 Specific Conversion Efficiency T10 Past elapsed time T20 Future elapsed time T200 Predicted Consumption Time Zone T300 Specific Time Slots V10 Past power consumption V20 Predicted Power Consumption

Claims

1. A recording unit that records the past power consumption of the load for each past elapsed time, A power consumption prediction unit creates predicted power consumption time periods by predicting the predicted power consumption of the load for each future elapsed time, based on the past power consumption for each past elapsed time recorded in the recording unit, A first calculation unit calculates the grid conversion efficiency, which is the power conversion efficiency for each future elapsed time, when the predicted amount of power consumption is supplied from the grid power source to the load at each future elapsed time during the predicted consumption time period. A second calculation unit calculates a specific conversion efficiency, which is the power conversion efficiency for each future elapsed time, when the predicted amount of power consumption is supplied to the load from a specific power source, which is provided separately from the grid power source, during the predicted consumption time period, at each future elapsed time; A special unit identifies the time period within the predicted consumption time period in which the specified conversion efficiency is equal to or greater than the system conversion efficiency as a specific time period, A planning unit that, in the specified time period, devises a supply plan indicating whether to supply power from the grid power source or the specified power source for each future elapsed time period in the predicted consumption time period, so that the predicted amount of power consumption is supplied to the load from the specified power source at each future elapsed time period in the predicted consumption time period, A power management device equipped with this device.

2. The first calculation unit calculates the grid conversion efficiency such that the grid conversion efficiency increases as the predicted power consumption increases. The power management device according to claim 1, wherein the second calculation unit calculates the specific conversion efficiency such that the specific conversion efficiency increases as the predicted power consumption increases.

3. When the predicted power consumption is less than a predetermined standard power consumption, the grid conversion efficiency becomes higher than the specified conversion efficiency. When the predicted power consumption is equal to or greater than the reference power consumption, the specific conversion efficiency becomes equal to or greater than the grid conversion efficiency. The power management device according to claim 1, wherein the identifying unit identifies the time period during which the predicted power consumption is equal to or greater than the standard power consumption within the predicted power consumption time period as the identified time period.

4. The aforementioned power consumption forecasting unit is A model generation unit generates a learning model that uses the past power consumption amounts for each past elapsed time as training data, takes the date and time for each past elapsed time as input, and outputs the past power consumption amounts for each past elapsed time. A machine learning unit inputs the date and time for each future elapsed time into the learning model during the predicted consumption time period and outputs the predicted power consumption for each future elapsed time. A power management device according to claim 1, having the following features.

5. The power management device according to claim 4, wherein the model generation unit adds the predicted power consumption for each future elapsed time output by the machine learning unit to the training data to generate the learning model.

6. A power management device as described in any one of claims 1 to 5, The aforementioned power supply system, The aforementioned specific power source, A power management system equipped with [specific features / equipment].

Images