Charge / discharge plan creation device and charge / discharge plan creation method

Abstract

To create a charge and discharge plan of a power storage battery with high accuracy, of which a profit by a power selling becomes the maximum.SOLUTION: A charge and discharge plan creation device that creates a charge and discharge plan of a power storage battery that charges a generated power of a power generation facility by performing a power generation by using a renewable energy, and discharges a power to be charged, contains: a charge and discharge plan creation part that calculates a charging amount in a plurality of charging frames performing a charging in the power storage battery and a discharge amount in a plurality of discharging frames performing a discharging from the power storage battery on the basis of a prediction value of a power price in each frame and a prediction value of the generated power of the power generation facility, constructed by dividing a day in each predetermined time; and a reduction part that reduces the discharge amount of each discharge frame so that a sum of the discharge amount in the plurality of discharge frames becomes a discharge available amount calculated on the basis of a sum of the generated power and an efficiency of the power storage battery.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to a charge-discharge plan creation device and a charge-discharge plan creation method for creating a charge-discharge plan for a storage battery.

Background Art

[0002] In recent years, with regard to the sale of electricity by operators of power generation facilities that generate electricity using renewable energy (such as solar power, wind power, hydropower, etc.), a Feed-in-Premium (FIP) system has been introduced. Along with the introduction of the FIP system, the above-mentioned power generation operators combine a power generation facility and a storage battery, charge the storage battery with the generated power during time periods such as daytime when the electricity price is relatively low, and discharge the power charged in the storage battery during time periods such as nighttime when the electricity price is relatively high, that is, a so-called time shift is performed, and a method for obtaining profits from the sale of electricity has been proposed. (See Patent Document 1)

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The amount of charge charged to the storage battery is the amount obtained by multiplying the amount of power to be charged to the storage battery by the charge efficiency α inherent to the storage battery. Further, the amount of discharge discharged from the storage battery is the amount obtained by multiplying the amount of charge of the storage battery by the discharge efficiency β inherent to the storage battery. In addition, in the storage battery, an inherent self-discharge amount γ of the storage battery occurs, in which the amount of charge decreases with the passage of time even though the storage battery is not discharging. Therefore, when comparing the amount of charge charged to the storage battery with the amount of discharge discharged from this charged amount of the storage battery, the amount of discharge becomes less than the amount of charge.

[0005] When performing the above time shift, a charge / discharge plan is created in advance to maximize the revenue generated from selling electricity on the target day of the time shift, and bids are made to sell electricity (discharge the battery) based on this charge / discharge plan. However, if the battery efficiency, such as the charging efficiency α, discharging efficiency β, and self-discharge amount γ, is not taken into consideration when creating the charge / discharge plan, the actual amount that can be discharged may be less than the amount of discharge planned in advance. As a result, it may be necessary to settle the imbalance for the shortfall in discharge amount, or to adjust for the shortfall in discharge amount in the advance market, which could make it difficult to maximize revenue on the target day.

[0006] This invention has been made in view of the above problems, and one of its objectives is to provide a charge / discharge plan creation device and a charge / discharge plan creation method that can create a highly accurate charge / discharge plan for a storage battery that maximizes revenue from electricity sales. [Means for solving the problem]

[0007] One of the present inventions for achieving the above objectives is a charge and discharge plan creation device for a storage battery that charges the power generated by a power generation facility that generates electricity using renewable energy and discharges the charged power, comprising: a charge and discharge plan creation unit that calculates the amount of charge in a plurality of charging units that charge the storage battery and the amount of discharge in a plurality of discharge units that discharge from the storage battery, based on the predicted value of the electricity price in each unit configured by dividing a day into predetermined time intervals and the predicted value of the power generated by the power generation facility; and a reduction unit that reduces the amount of discharge in each of the discharge units so that the sum of the discharge amounts in the plurality of discharge units becomes the amount that can be discharged based on the sum of the power generated and the efficiency of the storage battery.

[0008] According to the charge / discharge plan creation device of the present invention, it is possible to create a highly accurate charge / discharge plan for a storage battery that maximizes revenue from selling electricity.

[0009] Another aspect of the present invention for achieving the above objective is a charge / discharge planning device, wherein the reduction unit calculates the dischargeable amount based on the total generated power, the charging efficiency of the storage battery, the discharge efficiency, and the self-discharge amount.

[0010] According to the charge / discharge planning device of the present invention, since the factors used to calculate the dischargeable amount include the charging efficiency, discharging efficiency, and self-discharge amount of the storage battery, it is possible to create a highly accurate charge / discharge plan for the storage battery that maximizes revenue from selling electricity.

[0011] Another aspect of the present invention for achieving the above objective is a charge / discharge planning device in which the reduction unit reduces the discharge amount of each discharge unit, prioritizing the discharge unit of the discharge unit with a lower predicted electricity price.

[0012] According to the charge / discharge planning device of the present invention, the discharge amount is reduced in order of the discharge time slots with the lowest predicted electricity prices, making it possible to create a highly accurate charge / discharge plan for a storage battery that maximizes revenue from selling electricity.

[0013] Another aspect of the present invention for achieving the above objective is a charge / discharge plan creation device, wherein the charging time slot is a time slot included in a period when the predicted value of the electricity price on the day for which the charge / discharge plan is to be created is less than or equal to a first price, and the discharging time slot is a time slot included in a period when the predicted value of the electricity price on the day for which the plan is to be created is greater than or equal to a second price which is greater than or equal to the first price.

[0014] The charge-discharge plan created by the charge-discharge plan creation device of the present invention is a plan in which, on the day for which the charge-discharge plan of the storage battery is created, the storage battery is charged all at once during the daytime when the predicted value of electricity prices is relatively low, and the storage battery is discharged all at once during the nighttime when the predicted value of electricity prices is relatively high. As a result, the storage battery does not have to repeat the charge-discharge operation frequently, and it is possible to prevent the premature decline in charge efficiency and discharge efficiency.

[0015] Another aspect of the present invention for achieving the above objectives further includes a first correction unit in a charge / discharge plan creation device that corrects the values ​​of the charging efficiency, the discharge efficiency, and the self-discharge amount based on the difference between the sum of the actual discharge amounts of the plurality of discharge frames on the day for which the charge / discharge plan is to be created and the dischargeable amount.

[0016] According to the charge / discharge planning device of the present invention, the charging efficiency, discharging efficiency, and self-discharge amount are corrected according to the aging degradation status of the storage battery, making it possible to create a highly accurate charge / discharge plan for a storage battery that maximizes revenue from selling electricity.

[0017] Another aspect of the present invention for achieving the above objectives further includes, in a charge / discharge plan creation device, a second correction unit that corrects the predicted value of the electricity price on the day the charge / discharge plan is to be created based on the relationship between predicted and actual values ​​of electricity prices in the past, and a third correction unit that corrects the predicted value of the power generated by the power generation facility on the day the charge / discharge plan is to be created based on the relationship between predicted and actual values ​​of the power generated by the power generation facility in the past, wherein the charge / discharge plan creation unit calculates the charge amount and the discharge amount based on the corrected predicted value of the electricity price on the day the plan is to be created and the corrected predicted value of the power generated by the power generation facility on the day the charge / discharge plan is to be created.

[0018] According to the charge / discharge plan creation device of the present invention, the predicted value of the electricity price on the target date is corrected according to the predicted trend of electricity prices in the past, and the predicted value of the power generation of the power generation equipment on the target date is corrected according to the predicted trend of power generation of the power generation equipment in the past, thereby making it possible to create a highly accurate charge / discharge plan for a storage battery that maximizes revenue from selling electricity.

[0019] Another aspect of the present invention for achieving the above objective is a charge / discharge plan creation device, wherein the second correction unit corrects the predicted value of the electricity price on the day for which the charge / discharge plan is created, based on a first learning model generated by learning the relationship between predicted and actual values ​​of electricity prices in the past.

[0020] According to the charge / discharge plan creation device of the present invention, in order to correct the predicted value of the electricity price on the creation target date using the first learning model, it is possible to create a highly accurate charge / discharge plan for the storage battery that maximizes the profit from selling electricity.

[0021] Another aspect of the present invention for achieving the above object is that in the charge / discharge plan creation device, the third correction unit corrects the predicted value of the power generation of the power generation facility on the creation target date of the charge / discharge plan based on a second learning model generated by learning the relationship between the predicted value and the actual value of the power generation of the power generation facility in the past.

[0022] According to the charge / discharge plan creation device of the present invention, in order to correct the predicted value of the power generation of the power generation facility on the creation target date using the second learning model, it is possible to create a highly accurate charge / discharge plan for the storage battery that maximizes the profit from selling electricity.

[0023] In addition, the problems disclosed in the present application and the solutions thereto are clarified by the description in the section of the mode for carrying out the invention, the description in the drawings, and the like.

Effect of the Invention

[0024] According to the present invention, it is possible to create a highly accurate charge / discharge plan for the storage battery that maximizes the profit from selling electricity.

Brief Description of the Drawings

[0025] [Figure 1] It is a block diagram showing a schematic configuration of a power generation system 1 including a charge / discharge plan creation device 100. [Figure 2] It is a graph showing an example of the relationship between the predicted value of the electricity price, the predicted value of the power generation of the PV power generation facility 110, the planned value of the charge amount of the storage battery 120 in a plurality of charge slots, and the planned value of the discharge amount of the storage battery 120 in a plurality of discharge slots on the creation target date of the charge / discharge plan of the storage battery 120. [Figure 3] It is a block diagram showing the main functions of the charge / discharge plan creation device 100. [Figure 4] This figure shows an example of the first training data. [Figure 5] This figure shows an example of the second training data. [Figure 6] This figure shows the structure of a deep neural network, which is an example of the first learning model 1082A and the second learning model 1084A. [Figure 7] This is a flowchart showing the process when the charge / discharge planning device 100 trains the first learning model 1082A using the first learning data and trains the second learning model 1084A using the second learning data. [Figure 8A] This figure shows the planned discharge amounts of discharge frames D1 to D7 before the reduction unit 1050 performs the reduction process. [Figure 8B] This figure shows that the reduction unit 1050 performs a reduction process on the planned discharge amount of discharge unit D7, which has the lowest predicted power price after correction. [Figure 8C] This figure shows that the reduction unit 1050 performs a reduction process on the planned discharge amount of discharge unit D6, which is the next cheapest after discharge unit D7, based on the corrected predicted electricity price. [Figure 8D] This figure shows that the reduction unit 1050 performs a reduction process on the planned discharge amount of discharge unit D1, which is the next cheapest after discharge unit D6, based on the corrected predicted electricity price. [Figure 8E] This figure shows that the reduction unit 1050 performs a reduction process on the planned discharge amount of discharge unit D5, which has the next lowest predicted electricity price after discharge unit D1. [Figure 8F] This figure shows that the reduction unit 1050 performs a reduction process on the planned discharge amount of discharge unit D2, which is the next cheapest after discharge unit D5, based on the corrected predicted electricity price. [Figure 8G] This figure shows that the reduction unit 1050 performs a reduction process on the planned discharge amount of discharge piece D4, which has the next lowest predicted electricity price after discharge piece D2, based on the corrected price. [Figure 8H] This figure shows that the reduction unit 1050 performs a reduction process on the planned discharge amount of discharge piece D3, which has the next lowest (highest) predicted electricity price after the correction, after the discharge piece D4. [Figure 9] This block diagram shows an example of the hardware of the information processing device 10 that realizes the functions of the charge / discharge planning device 100 shown in Figure 3. [Figure 10] This flowchart shows an example of the process when the charge / discharge plan creation device 100 creates a charge / discharge plan for the battery 120 on the designated day for creating the charge / discharge plan. [Modes for carrying out the invention]

[0026] The following matters will become clear from this specification and the accompanying drawings. The present invention will be described below with reference to the accompanying drawings, with reference to one embodiment thereof.

[0027] Figure 1 is a block diagram showing the schematic configuration of a power generation system 1 including a charge / discharge planning device 100 according to this embodiment. In this embodiment, the power generation equipment that generates electricity using renewable energy is, for example, a photovoltaic power generation equipment (hereinafter referred to as "PV power generation equipment" (PV: Photo Voltaic)) 110 that generates electricity using sunlight.

[0028] The power generation system 1 comprises a charge / discharge planning device 100, a PV power generation equipment 110, a storage battery 120, a storage battery control device 130, a weather information provision device 150, an electricity price forecasting device 160, and a power generation forecasting device 170.

[0029] The charge / discharge planning device 100, PV power generation equipment 110, storage battery 120, storage battery control device 130, weather information provision device 150, electricity price forecasting device 160, and power generation forecasting device 170 are connected in a manner that enables communication via a communication network 180. The communication network 180 can be, for example, a LAN (Local Area Network), WAN (Wide Area Network), dedicated line, power line communication network, or various public communication networks. In addition to the communication network 180, the PV power generation equipment 110 and storage battery 120 are also connected to a power grid 190 operated by a general power transmission and distribution company, etc.

[0030] The PV power generation equipment 110 includes a photovoltaic power generation panel (not shown) composed of, for example, polycrystalline silicon power generation elements, monocrystalline silicon power generation elements, thin-film power generation elements, etc. The PV power generation equipment 110 also includes a power conditioner (PCS: Power Conditioning Subsystem) 110A that converts the power generated by the photovoltaic power generation panel from DC to AC and supplies it to the power grid 190. The power generated by the PV power generation equipment 110 is either supplied to the power grid 190 via the power conditioner 110A (sold to a general transmission and distribution company) or supplied to the storage battery 120 (charged by the storage battery 120). In this embodiment, the power generated by the PV power generation equipment 110 is supplied to the storage battery 120 during multiple charging periods included in the charging time.

[0031] The battery 120 is, for example, a lead-acid battery, a lithium-ion battery, a sodium-sulfur battery, a nickel-metal hydride battery, a redox flow battery, a fuel cell, a capacitor battery, etc. The battery 120 is equipped with a power conditioner (PCS) 120A for charging and discharging in and out of the power grid 190.

[0032] The weather information provider 150 provides the electricity price forecasting device 160 and the power generation forecasting device 170 with weather information (e.g., temperature, solar radiation, etc.) for the area where the PV power generation equipment 110 is installed on the day for which the charge / discharge plan for the storage battery 120 is to be created. The weather information provider 150 may obtain the above weather information from a weather observation station in the area where the PV power generation equipment 110 is installed, or it may obtain the above weather information from a weather information server (not shown) managed by the Japan Meteorological Agency.

[0033] The electricity price prediction device 160 predicts the electricity price (yen / kW) of electricity traded on the target date for creating the charge / discharge plan for the battery 120, by referring to weather information obtained from the weather information provider 150, as well as information such as the date, time, and day of the week. In this embodiment, the electricity price prediction device 160 predicts the electricity price for each time frame, for example, with a 30-minute time frame as one frame. Furthermore, as a method for predicting electricity prices, the electricity price prediction device 160 can use, for example, a well-known technique that predicts electricity prices by combining the above-mentioned weather information, date, time, and day of the week with analytical techniques such as artificial intelligence. In addition, the electricity price prediction device 160 may obtain predicted electricity prices published by a third party via the communication network 180 or the internet.

[0034] The power generation forecasting device 170 predicts the power generation (MWh) for each cycle of the PV power generation equipment 110 based on information such as weather information obtained from the weather information provider 150 for the days for which the charge / discharge plan of the storage battery 120 is to be created, location information (latitude, longitude, etc.) of the place where the PV power generation equipment 110 is installed, the panel area of ​​the PV power generation equipment 110's solar panels that can receive sunlight, and the tilt angle at which the solar panels rise from the ground so that the PV power generation equipment 110 can generate power efficiently. As a method for the power generation forecasting device 170 to predict the power generation of the PV power generation equipment 110, for example, a physical model that combines analytical techniques to predict power generation with the above-mentioned weather information, location information of the PV power generation equipment 110, panel area and tilt angle of the solar panels, etc., or a statistical model that combines analytical techniques such as artificial intelligence that has learned the relationship between past weather information and actual power generation values.

[0035] The charge / discharge planning device 100 obtains predicted electricity prices from the electricity price forecasting device 160 and predicted power generation values ​​from the power generation forecasting device 170 for the days for which a charge / discharge plan for the storage battery 120 is to be created. Based on the predicted electricity prices and power generation values, the device adjusts the charge amounts of multiple charging frames included in time periods when the predicted electricity prices are relatively low and the predicted power generation values ​​of the power generation facilities 110 are relatively high (charging time periods suitable for charging the storage battery 120), and the discharge amounts of multiple discharge frames included in time periods when the predicted electricity prices are relatively high and the predicted power generation values ​​of the power generation facilities 110 are relatively low (discharge time periods suitable for discharging the storage battery 120). The device then creates a charge / discharge plan according to the charge amounts of multiple charging frames and discharge amounts of multiple discharge frames that maximize the revenue from selling electricity through the discharge of the storage battery 120. Details of the charge / discharge planning device 100 will be described later.

[0036] The battery control device 130 acquires the charge and discharge plan finally determined by the charge and discharge plan creation device 100, and controls the charge and discharge operation of the battery 120 so that the battery 120 charges according to the charge amount for each of the multiple charge frames planned in the charge and discharge plan, and the battery 120 discharges according to the discharge amount for each of the multiple discharge frames planned in the charge and discharge plan.

[0037] Figure 2 is a graph showing an example of the relationship between the predicted electricity price by the electricity price forecasting device 160, the predicted power generation of the PV power generation facility 110 by the power generation power forecasting device 170, the planned charge amount of the battery 120 in multiple charging frames, and the planned discharge amount of the battery 120 in multiple discharge frames, on the day for which the charge and discharge plan of the battery 120 is to be created. In particular, Figure 2 is a graph showing the initial charge and discharge plan (initial charge and discharge plan value) created by the charge and discharge plan creation device 100 so as to maximize the revenue from selling electricity by discharging the battery 120 on the day for which the charge and discharge plan of the battery 120 is to be created. Note that the predicted electricity price is a predicted value corrected by the electricity price correction unit 1020 described later, and the predicted power generation of the PV power generation facility 110 is a predicted value corrected by the power generation correction unit 1030 described later.

[0038] In Figure 2, the horizontal axis represents the time in a day for which the charge / discharge plan for the battery 120 is to be created, and shows 48 time units of 30 minutes each. Also in Figure 2, the left vertical axis scale shows the power generated by the PV power generation equipment 110 in units of 0.5 (MWh), and the right vertical axis scale shows the electricity price in units of 5 (yen / kW) in the market where electricity is bought and sold. Furthermore, in Figure 2, the dashed line shows the predicted value of the electricity price on the day for which the charge / discharge plan for the battery 120 is to be created, the double dashed line shows the predicted value of the power generated by the PV power generation equipment 110 on the day for which the charge / discharge plan for the battery 120 is to be created, the black vertical line shows the planned value of the charge amount of the battery 120 in 10 charging units C1 to C10 included in the charging time period, and the white vertical line shows the planned value of the discharge amount of the battery 120 in 7 discharge units D1 to D7 included in the discharge time period. In this embodiment, the charging time period is from 9:30 to 14:00 when the predicted electricity price is, for example, 19.85 (yen (first price) / kW) or less, and the discharge time period is from 16:00 to 19:00 when the predicted electricity price is, for example, 23.02 (yen (second price) / kW) or more. The charging time period includes 10 charging frames C1 to C10 by dividing the period from 9:30 to 14:00 into 30-minute units, and the discharge time period includes 7 discharge frames D1 to D7 by dividing the period from 16:00 to 19:00 into 30-minute units.

[0039] Here, when the battery 120 is charged with the power generated by the PV power generation equipment 110, the planned value of the charge amount per unit is equal to the capacity of the power conditioner 110A / 2 when the predicted value of the power generated per unit of the PV power generation equipment 110 is 2 or more than the capacity of the power conditioner 110A, and on the other hand, when the predicted value of the power generated per unit of the PV power generation equipment 110 is less than the capacity of the power conditioner 110A, it is equal to the predicted value of the power generated per unit of the PV power generation equipment 110. Note that the capacity of the power conditioner 110A / 2 is the capacity that the power conditioner 110A can handle for one unit of the power generated by the PV power generation equipment 110. In this embodiment, the predicted value of the power generated by the PV power generation equipment 110 during the charging period is set to be 2 or more than the capacity of the power conditioner 110A. In other words, the maximum planned charge amount of the battery 120 in charging frames C1 to C10 is limited to the capacity of the power conditioner 110A / 2 (for example, 2.5 (MWh)). Similarly, the maximum planned discharge amount of the battery 120 in discharging frames D1 to D7 is also limited to the capacity of the power conditioner 110A / 2.

[0040] In Figure 2, the total planned discharge amount of the battery 120 during the discharge period is set to be less than the total planned charge amount of the battery 120 during the charging period. However, the settings are not limited to the example in Figure 2. For example, the settings may be made so that the total planned charge amount of the battery 120 during the charging period is equal to the total planned discharge amount of the battery 120 during the discharge period, so that the State of Charge (SOC) of the battery 120 before the charging period and after the discharge period are both at their lower limits.

[0041] The predicted electricity prices during the charging period increase in the order of charging time slots C6, C7, C8, C1, C5, C4, C3, C2, C9, and C10. Therefore, the planned charge amounts are set in the order of charging time slots C6, C7, C8, C1, C5, C4, C3, C2, C9, and C10. Similarly, the predicted electricity prices during the discharging period decrease in the order of discharging time slots D3, D4, D2, D5, D1, D6, and D7. Therefore, the planned discharge amounts are set in the order of discharging time slots D3, D4, D2, D5, D1, D6, and D7.

[0042] In Figure 2, the charging and discharging schedule for battery 120 is created, with the discharging period set after the charging period. However, if the predicted electricity price rises to 23.02 yen / kW or higher during the charging period, or falls to 19.85 yen / kW or lower during the discharging period, a discharging period may be inserted within the charging period, or a charging period may be inserted within the discharging period. Also, in Figure 2, different predicted electricity prices (19.85 yen / kW) for setting the charging period and 23.02 yen / kW) for setting the discharging period are selected, but this is not limited to these values. For example, one predicted electricity price could be selected to separate the charging and discharging periods, with the period when the electricity price is below this predicted value set as the charging period, and the period when the electricity price is above this predicted value set as the discharging period.

[0043] The charge / discharge plan creation device 100 creates the initial charge / discharge plan values ​​shown in Figure 2, which include a charge plan in which the planned charge amounts of the battery 120 are set in charge frames C1 to C10, and a discharge plan in which the planned discharge amounts of the battery 120 are set in discharge frames D1 to D7. Note that the initial charge / discharge plan values ​​may also be input to the charge / discharge plan creation device 100 from an external device (not shown) capable of creating these initial charge / discharge plan values.

[0044] Figure 3 is a block diagram showing the main functions of the charge / discharge planning device 100.

[0045] The charge / discharge plan creation device 100 is composed of an input unit 1010, a power price correction unit 1020 (second correction unit), a power generation correction unit 1030 (third correction unit), a charge / discharge plan creation unit 1040, a reduction unit 1050, a determination unit 1060, a charge / discharge plan determination unit 1070, a storage unit 1080, and an output unit 1090.

[0046] The memory unit 1080 is configured to include the first memory area 1081 to the ninth memory area 1089.

[0047] The first memory area 1081 stores a control program for operating the charge / discharge planning device 100.

[0048] The second memory area 1082 stores a first learning model 1082A that corrects the predicted value of the electricity price on the day for which the charge-discharge plan for the battery 120 is created, taking into account the discrepancy between the predicted and actual values ​​of the electricity price on predetermined past days (for example, multiple days in the same season or month in the past), in order to maximize the revenue from selling electricity when the charge-discharge plan for the battery 120 is executed, that is, to improve the accuracy of the charge-discharge plan on the day for which the plan is created.

[0049] The third memory area 1083 stores the first training data for the first learning model 1082A to perform machine learning.

[0050] Figures 4 and 5 show examples of the first and second training data, respectively.

[0051] The first training data is data that associates the first feature with the first label. The first feature includes, for example, weather information (temperature, atmospheric pressure, humidity, solar radiation, etc.) predicted for the installation area of ​​the PV power generation facility 110 on a predetermined date in the past, information on the date and day of the week indicating the predetermined date in the past, a predicted value of the electricity price (yen / kW) predicted for electricity buying and selling transactions on the predetermined date in the past, and information on the actual value of the electricity price (yen / kW) determined on the predetermined date in the past. On the other hand, the first label is corrected information obtained by the party creating the first training data, which corrects the predicted value of the electricity price so that the discrepancy between the predicted value and the actual value of the electricity price is reduced, taking into account the weather information, date and day of the week on the predetermined date in the past. The first training data is data that is accumulated by associating the first feature and the first label for each of multiple past days with the same day.

[0052] The first learning model 1082A takes in the first learning data and learns in advance the correction trends of predicted electricity prices for the same season or month in the past. When the first learning model 1082A receives information on weather, date and time, and day of the week, as well as predicted electricity prices, for the day on which the charge and discharge plan for the battery 120 is to be created, it corrects the predicted electricity price for the day on which the charge and discharge plan for the battery 120 is to be created, based on the correction trends of predicted electricity prices for multiple days with similar weather in the same season or month in the past, so as to maximize the revenue from selling electricity when the charge and discharge plan for the battery 120 is executed, and outputs data showing this corrected predicted electricity price.

[0053] The fourth memory area 1084 stores a second learning model 1084A that corrects the predicted value of the power generated on the day for which the charge-discharge plan for the battery 120 is created, frame by frame, taking into account the discrepancy between the predicted and actual values ​​of the power generated by the PV power generation equipment 110 on predetermined past days (for example, multiple days in the same season or month in the past), in order to maximize the revenue from selling electricity when the charge-discharge plan for the battery 120 is executed, that is, to improve the accuracy of the charge-discharge plan on the day for which the plan is created.

[0054] The fifth memory area 1085 stores the second training data for the second learning model 1084A to perform machine learning.

[0055] The second training data is data that associates the second feature with the second label. The second feature includes, for example, weather information predicted for the installation area of ​​the PV power generation facility 110 on a predetermined date in the past, date and time information indicating the predetermined date in the past, predicted value (MWh) of the power generated by the PV power generation facility 110 on the predetermined date in the past, and actual value (MWh) of the power generated by the PV power generation facility 110 on the predetermined date in the past. On the other hand, the second label is corrected information in which the party creating the second training data corrects the predicted value of the power generated so that the discrepancy between the predicted value and the actual value of the power generated by the PV power generation facility 110 is reduced, taking into account the date and time indicating the predetermined date in the past and the weather information at that time. The second training data is data that is accumulated by associating the second feature and the second label for each of multiple past days with the same day.

[0056] The second learning model 1084A takes in the second learning data and pre-learns the correction trends for predicted power generation values ​​of the PV power generation facility 110 in the same season or month in the past. When the second learning model 1084A receives information on the weather forecast, date and time information for the installation area of ​​the PV power generation facility 110, and predicted power generation value of the PV power generation facility 110 for the day on which the charge and discharge plan for the battery 120 is to be created, it corrects the predicted power generation value of the PV power generation facility 110 for the day on which the charge and discharge plan for the battery 120 is to be created, based on the correction trends for predicted power generation values ​​for multiple days with similar weather conditions in the same season or month in the past, so as to maximize the revenue from selling electricity when the charge and discharge plan for the battery 120 is executed, and outputs data showing this corrected predicted power generation value.

[0057] Furthermore, since the first and second learning data are only necessary when training the first learning model 1082A and the second learning model 1084A, they may be stored in a memory unit of an external computer (not shown) instead of being stored in the memory unit 1080. The charge / discharge plan creation device 100 may then acquire the first and second learning data from the memory unit of the external computer via the communication network 180, and use the first and second learning data to train the first learning model 1082A and the second learning model 1084A on the charge / discharge plan creation device 100. Alternatively, the first learning model 1082A and the second learning model 1084A may be trained using the first and second learning data in an external computer, and the charge / discharge plan creation device 100 may acquire the trained first learning model 1082A and the second learning model 1084A from the external computer via the communication network 180 and store them in the second storage area 1082 and the fourth storage area 1084 of the storage unit 1080.

[0058] The first learning model 1082A and the second learning model 1084A can make more accurate predictions as they are trained using more of the first and second training data, respectively. While the first learning model 1082A and the second learning model 1084A are assumed to be, for example, deep neural networks (DNNs), they may also be other types of models such as gradient boosting (GBDT) or decision trees.

[0059] Figure 6 shows the structure of a deep neural network, which is an example of the first learning model 1082A and the second learning model 1084A.

[0060] The first learning model 1082A and the second learning model 1084A each consist of three layers: an input layer 2010, an intermediate layer 2020, and an output layer 2030.

[0061] In the case of the first learning model 1082A, the input layer 2010 is a layer into which data of the same type as the first feature of the first learning data is input. Specifically, the input layer 2010 is input with information on the date and day of the week for the day on which the charge and discharge plan for the battery 120 is to be created, weather information predicted for the area where the PV power generation equipment 110 is installed, and data showing the predicted value of the electricity price. The hidden layer 2020 is a layer that includes one or more hidden layers consisting of one or more nodes containing parameters that are adjusted by learning using the first learning data. The hidden layer 2020 corrects the predicted value of the electricity price for the day on which the charge and discharge plan for the battery 120 is to be created based on the data input to the input layer 2010. The output layer 2030 is a layer that outputs data showing the corrected predicted value of the electricity price processed in the hidden layer 2020.

[0062] In the case of the second learning model 1084A, the input layer 2010 is a layer into which data of the same type as the second feature of the second learning data is input. Specifically, the input layer 2010 is input with date and time information for the day for which the charge and discharge plan for the battery 120 is to be created, weather information predicted for the installation area of ​​the PV power generation facility 110, and data showing the predicted value of the power generated by the PV power generation facility 110. The hidden layer 2020 is a layer that includes one or more hidden layers consisting of one or more nodes containing parameters that are adjusted by learning using the second learning data. The hidden layer 2020 corrects the predicted value of the power generated by the PV power generation facility 110 for the day for which the charge and discharge plan for the battery 120 is to be created based on the data input to the input layer 2010. The output layer 2030 is a layer that outputs data showing the corrected predicted value of the power generated after processing in the hidden layer 2020.

[0063] Figure 7 is a flowchart showing the process when the charge / discharge planning device 100 learns the first learning model 1082A using the first learning data and learns the second learning model 1084A using the second learning data.

[0064] When training the first learning model 1082A, the charge / discharge planning device 100 generates first learning data that associates, for each day, the following: information on the date and day of the week for each of the past days, weather information predicted for the area where the PV power generation equipment 110 is installed, first feature quantities which are explanatory variables indicating predicted and actual values ​​of electricity prices, and first labels which are dependent variables indicating the predicted value of the corrected electricity price (S3010).

[0065] Next, the charge / discharge planning device 100 uses the first learning data to train the first learning model 1082A (S3020).

[0066] On the other hand, when training the second learning model 1084A, the charge / discharge planning device 100 generates second learning data that associates information on the date and time for each of the past multiple days, weather information predicted for the area where the PV power generation facility 110 is installed, a second feature which is an explanatory variable indicating the predicted and actual values ​​of the power generated by the PV power generation facility 110, and a second label which is an objective variable indicating the predicted value of the power generated after correction (S3010).

[0067] Next, the charge / discharge planning device 100 uses the second learning data to train the second learning model 1084A (S3020).

[0068] Furthermore, the charge / discharge planning device 100 may also perform a verification of the prediction accuracy of the trained first learning model 1082A and second learning model 1084A. In that case, the charge / discharge planning device 100 prepares training data and verification data for the first learning data and the second learning data, trains the first learning model 1082A and the second learning model 1084A using the training data, and verifies the first learning model 1082A and the second learning model 1084A using the verification data.

[0069] Returning to Figure 3, the sixth memory area 1086 stores data for one day, corresponding to the same time slot on the same day, including predicted electricity prices for 48 time slots per day obtained from the electricity price forecasting device 160, and actual electricity prices for the same 48 time slots per day. This data is accumulated and stored for the number of past days in which electricity trading has taken place. Furthermore, in addition to the data showing the predicted and actual electricity prices already stored in the sixth memory area 1086, data for one day at a time, corresponding to the same time slot on the same day, including predicted electricity prices for 48 time slots per day from the present onward, and actual electricity prices for the same 48 time slots per day, obtained from the electricity price forecasting device 160. This data showing the predicted and actual electricity prices is used as the first feature of the first training data stored in the third memory area 1083. In particular, new data showing predicted and actual electricity prices added to the sixth memory area 1086 are added to the first feature of the first training data, and new data corrected from this predicted electricity price are added to the first label of the first training data. The first training model 1082A improves its accuracy in grasping the correction trend of predicted electricity prices for the same season or month in the past by training using the updated first training data. The timing of retraining of the first training model 1082A may be each time the first feature and first label constituting the first training data are updated, each time the first feature and first label constituting the first training data are updated a certain number of times, or at regular intervals. The actual electricity price is, for example, the contract price when bidding on the buying and selling of electricity in the electricity trading market. Data showing this contract price is obtained from an external computer or server via the input unit 1010 and then associated with the predicted electricity price.

[0070] The seventh memory area 1087 stores data for a full day, where the predicted power generation values ​​for 48 time slots per day, obtained from the power generation forecasting device 170, and the actual power generation values ​​for the same 48 time slots per day are associated with the same time slot on the same day. This data is accumulated for the number of past days in which power buying and selling transactions have taken place. Furthermore, in addition to the data showing the predicted and actual power generation values ​​already stored in the seventh memory area 1087, data for each day, where the predicted power generation values ​​for 48 time slots per day and the actual power generation values ​​for the same 48 time slots per day, obtained from the power generation forecasting device 170, are associated with the same time slot on the same day, and are sequentially accumulated and stored. This data showing the predicted and actual power generation values ​​is used as the second feature of the second training data stored in the fifth memory area 1085. In particular, new data showing predicted and actual values ​​of power generation added to the seventh memory area 1087 are added to the second feature of the second training data, and new data corrected from this predicted power generation value is added to the second label of the second training data. The second training model 1084A improves its accuracy in grasping the correction trend of predicted power generation values ​​for the same season or month in the past by training using the updated second training data. The timing of retraining of the second training model 1084A may be each time the second feature and second label constituting the second training data are updated, each time the second feature and second label constituting the second training data are updated a certain number of times, or at regular intervals. Data showing the actual power generation value of the PV power generation facility 110 is acquired from an external computer or server via the input unit 1010 and then associated with the predicted power generation value.

[0071] The eighth memory area 1088 stores data showing the initial values ​​of the charge / discharge plan for the battery 120, as shown in Figure 2, which was initially created by the charge / discharge plan creation unit 1040.

[0072] The ninth memory area 1089 stores the updated charge / discharge plan value, which is the initial value of the charge / discharge plan updated to maximize the revenue from selling electricity on the day for which the charge / discharge plan of the battery 120 is created. Details of the updated charge / discharge plan value will be described later.

[0073] When the charge / discharge plan creation device 100 is determined to create a charge / discharge plan for the battery 120, the input unit 1010 obtains a predicted value for the electricity price on the day for which the charge / discharge plan for the battery 120 is to be created from the electricity price prediction device 160, and also obtains a predicted value for the power generated by the PV power generation equipment 110 on the day for which the charge / discharge plan for the battery 120 is to be created from the power generation prediction device 170.

[0074] The power price correction unit 1020, upon receiving data in the input unit 1010 indicating the predicted power price for the day for which the charge / discharge plan for the battery 120 is to be created, instructs the first learning model 1082A to correct this predicted power price so as to maximize the revenue when selling electricity according to the charge / discharge plan for the battery 120. The power price correction unit 1020 also provides the charge / discharge plan creation unit 1040 with the corrected predicted power price output from the first learning model 1082A.

[0075] The power generation correction unit 1030, upon receiving data from the input unit 1010 indicating the predicted power generation of the PV power generation facility 110 on the day for which the charge / discharge plan for the battery 120 is to be created, instructs the second learning model 1084A to correct this predicted power generation value so as to maximize the revenue when electricity is sold according to the charge / discharge plan for the battery 120. The power generation correction unit 1030 also provides the charge / discharge plan creation unit 1040 with the corrected predicted power generation value output from the second learning model 1084A.

[0076] The charge / discharge plan creation unit 1040 uses the corrected predicted power price output from the first learning model 1082A and the corrected predicted power generation value output from the second learning model 1084A to create an initial charge / discharge plan value, for example, as shown in Figure 2. This initial charge / discharge plan value is stored in the eighth memory area 1088 of the memory unit 1080.

[0077] In the initial charge / discharge plan, the planned charge amount of the battery 120 planned for multiple charging frames included in the charging time period does not take into account the charging efficiency α specific to the battery 120, and the planned discharge amount of the battery 120 planned for multiple discharge frames included in the discharging time period does not take into account the discharge efficiency β and self-discharge amount γ specific to the battery 120. Therefore, the reduction unit 1050 reduces the difference between the sum of the uncorrected planned discharge amounts and the dischargeable amount, starting with the discharge frames with the lowest predicted power prices after correction, until the sum of the uncorrected planned discharge amounts planned for multiple discharge frames becomes the dischargeable amount that takes into account the charging efficiency α, discharge efficiency β, and self-discharge amount γ of the battery 120.

[0078] The reduction unit 1050 reduces the discharge amount of each discharge frame D1 to D7 so that the sum of the planned values ​​of the discharge amounts before correction in each discharge frame D1 to D7 becomes the dischargeable amount (planned value of the corrected discharge amount) calculated based on the total power generated by the PV power generation equipment 110 in each discharge frame D1 to D7 and the efficiency of the storage battery 120 (charging efficiency α, discharge efficiency β, self-discharge amount γ). The reduction unit 1050 calculates the dischargeable amount when performing the discharge amount reduction process for each discharge frame D1 to D7.

[0079] Figures 8A to 8H are graphs illustrating an example of the process by which the reduction unit 1050 reduces the sum of the planned discharge amounts of multiple discharge frames included in the discharge time period until it becomes the dischargeable amount.

[0080] Figures 8A to 8H are graphs showing only the planned discharge amounts for discharge frames D1 to D7 included in the discharge time period of Figure 2. The horizontal axis represents the time for discharge frames D1 to D7 during the discharge time period from 16:00 to 19:00, and the vertical axis scale represents the power generated in units of 0.5 (MWh) at the PV power generation facility 110. The corrected predicted power prices for discharge frames D1 to D7 increase in the order of D7, D6, D1, D5, D2, D4, and D3.

[0081] Specifically, Figure 8A shows the planned discharge amounts of discharge units D1 to D7 before the reduction unit 1050 performs the reduction process. Figure 8B shows that the reduction unit 1050 performs the reduction process on the planned discharge amount before correction for discharge unit D7, which has the lowest predicted corrected electricity price. Figure 8C shows that the reduction unit 1050 performs the reduction process on the planned discharge amount before correction for discharge unit D6, which has the next lowest predicted corrected electricity price after discharge unit D7. Figure 8D shows that the reduction unit 1050 performs the reduction process on the planned discharge amount before correction for discharge unit D1, which has the next lowest predicted corrected electricity price after discharge unit D6. Figure 8E shows that the reduction unit 1050 performs the reduction process on the planned discharge amount before correction for discharge unit D5, which has the next lowest predicted corrected electricity price after discharge unit D1. Furthermore, Figure 8F shows that the reduction unit 1050 performs a reduction process on the uncorrected planned discharge amount of discharge piece D2, which has the next lowest predicted power price after correction after discharge piece D5. Furthermore, Figure 8G shows that the reduction unit 1050 performs a reduction process on the uncorrected planned discharge amount of discharge piece D4, which has the next lowest predicted power price after correction after discharge piece D2. Furthermore, Figure 8H shows that the reduction unit 1050 performs a reduction process on the uncorrected planned discharge amount of discharge piece D3, which has the next lowest (highest) predicted power price after correction after discharge piece D4. The dashed lines in Figures 8B to 8H show the reduction values ​​obtained by reducing the uncorrected planned discharge amount.

[0082] As is clear from Figure 2, the planned charge values ​​for charging frames C1 to C9 are 2.5 (MWh) each, the planned charge value for charging frame C10 is 1 (MWh), the planned uncorrected discharge values ​​for discharging frames D1 to D6 are 2.5 (MWh) each, and the planned uncorrected discharge value for discharging frame D7 is 2 (MWh). Also, for the sake of explanation, we assume that the charging efficiency α = 0.85, the discharging efficiency β = 0.85, and the self-discharge amount γ per frame for discharging frames D1 to D7 = 0.1 (MWh).

[0083] In Figure 2, the total planned charge amount for battery 120 in charging frames C1 to C10 included in the charging time period is 2.5 × 9 + 1 = 23.5 (MWh). On the other hand, the actual chargeable amount that can be charged to battery 120, considering the charging efficiency α, is 23.5 × 0.85 = 19.975 (MWh). The difference between the total planned charge amount for battery 120, 23.5 (MWh), and the actual chargeable amount of battery 120, 19.975 (MWh), is 3.525 (MWh). This difference of 3.525 (MWh), considering the charging efficiency α, will not be charged to battery 120, but it is included in the planned charge amount. Therefore, among the discharge frames D1 to D7 included in the discharge time period, in order of the lowest predicted corrected electricity price, it is necessary to reduce the planned discharge amount by the difference of 3.525 (MWh) until it becomes 0 (MWh).

[0084] The following describes an example of the weight reduction operation of the weight reduction unit 1050, with reference to Figures 8A to 8H.

[0085] The depletion unit 1050 refers to the initial charge / discharge plan values ​​stored in the eighth memory area 1088 and calculates the above difference value of 3.525 (MWh).

[0086] Then, the reduction unit 1050 starts a reduction process that takes into account the charging efficiency α, discharge efficiency β, and self-discharge amount γ, starting from the planned value of the discharge amount of the storage battery 120 before correction in discharge frames D1 to D7 shown in Figure 8A.

[0087] <Figure 8B> The reduction unit 1050 calculates a discharge amount of 1.7 (MWh) in discharge frame D7 that takes into account the discharge efficiency β by multiplying the planned value 2 (MWh) of the uncorrected discharge amount of the storage battery 120 in discharge frame D7, where the corrected predicted electricity price is the lowest, by the discharge efficiency β.

[0088] Next, the reduction unit 1050 reduces the discharge amount in discharge frame D7 from the planned value of the discharge amount before correction, taking into account the discharge efficiency β, and calculates a reduction value of 0.3 (MWh) that takes into account the discharge efficiency β.

[0089] Next, the reduction unit 1050 subtracts the reduction value considering the discharge efficiency β and the self-discharge amount γ from the planned value of the discharge amount before correction in the discharge ring D7 to calculate the dischargeable amount of 1.6 (MWh) in the discharge ring D7.

[0090] Next, the reduction unit 1050 reduces the dischargeable amount in the discharge ring D7 by 1.6 (MWh) of the above difference of 3.525 (MWh).

[0091] As a result, the corrected planned discharge amount in discharge choke D7 changes from 2 (MWh) to 0 (MWh).

[0092] <Figure 8C> Next, the reduction unit 1050 calculates a discharge amount of 2.125 (MWh) in discharge frame D6, which has the next lowest predicted electricity price after discharge frame D7, by multiplying the pre-correction discharge amount of the battery 120, which is 2.5 (MWh), by the discharge efficiency β, to determine the discharge amount of 2.125 (MWh) in discharge frame D6 that takes the discharge efficiency β into account.

[0093] Next, the reduction unit 1050 reduces the discharge amount in discharge choke D6 from the planned value of the discharge amount before correction, taking into account the discharge efficiency β, and calculates a reduction value of 0.375 (MWh) that takes into account the discharge efficiency β.

[0094] Next, the reduction unit 1050 reduces the discharge amount in discharge piece D6 by the planned value of the discharge amount before correction, by the reduction value considering the discharge efficiency β and the self-discharge amount γ, and calculates the dischargeable amount in discharge piece D6 of 2.025 (MWh).

[0095] Next, the reduction unit 1050 reduces the dischargeable amount in discharge piece D6 by 1.925 (MWh), which is the amount of the difference of 3.525 (MWh) that could not be reduced from the dischargeable amount in discharge piece D7. At this point, the reduction process of the difference of 3.525 (MWh) is completed.

[0096] As a result, the corrected planned discharge amount in discharge choke D6 changes from 2.5 (MWh) to 0.1 (MWh).

[0097] <Figure 8D> Next, the reduction unit 1050 calculates a discharge amount of 2.125 (MWh) in discharge frame D1, which has the next lowest predicted electricity price after the corrected price, by multiplying the planned discharge amount of the battery 120 before correction, which is 2.5 (MWh), by the discharge efficiency β, to determine the discharge amount of 2.125 (MWh) in discharge frame D1 that takes the discharge efficiency β into account.

[0098] Next, the reduction unit 1050 reduces the discharge amount in discharge frame D1 from the planned value of the discharge amount before correction, taking into account the discharge efficiency β, and calculates a reduction value of 0.375 (MWh) that takes into account the discharge efficiency β.

[0099] Next, the reduction unit 1050 reduces the discharge amount in discharge piece D1 by the planned value of the discharge amount before correction, by the reduction value considering the discharge efficiency β and the self-discharge amount γ, and calculates the dischargeable amount in discharge piece D1 to be 2.025 (MWh).

[0100] As a result, the corrected planned discharge amount in discharge block D1 changes from 2.5 (MWh) to 2.025 (MWh).

[0101] <Figures 8E to 8H> Next, the reduction unit 1050 performs the same process as the reduction treatment in the reduction unit 1050 shown in Figure 8D. As a result, the corrected planned discharge amounts in discharge pieces D5, D2, D4, and D3 change from 2.5 (MWh) to 2.025 (MWh).

[0102] In Figure 8H, the sum of the corrected planned discharge amounts shown by the solid lines for discharge frames D1 to D6 represents the dischargeable amount of battery 120 during the discharge period, and the sum of the reduced discharge amounts shown by the dashed lines for discharge frames D1 to D7 represents the difference between the sum of the uncorrected discharge amounts shown by the solid lines for discharge frames D1 to D7 in Figure 8A and the dischargeable amount shown in Figure 8H. Then, in the initial charge / discharge plan values ​​shown in Figure 2, the uncorrected planned discharge amounts for discharge frames D1 to D7 during the discharge period are replaced with the corrected planned discharge amounts shown in Figure 8H to obtain the updated charge / discharge plan values.

[0103] In this way, the reduction unit 1050 performs a reduction process on the planned values ​​of the uncorrected discharge amounts for all discharge frames D1 to D7 included in the discharge time period, taking into account the charging efficiency α, the discharge efficiency β, and the self-discharge amount γ, and then completes the series of reduction processes.

[0104] Furthermore, the reduction unit 1050 has pre-set values ​​for charging efficiency α, discharging efficiency β, and self-discharge amount γ used when performing the reduction process. The reduction unit 1050 functions as a first correction unit that corrects the values ​​of charging efficiency α, discharging efficiency β, and self-discharge amount γ based on the difference between the sum of the actual discharge amounts of discharge frames D1 to D7 on the day for which the charge / discharge plan is created, and the planned discharge amount, which is the corrected discharge amount of discharge frames D1 to D7. These values ​​of charging efficiency α, discharging efficiency β, and self-discharge amount γ are corrected by the reduction unit 1050 based on the degree of deviation between the sum of the actual discharge amounts of discharge frames D1 to D7 and the discharge amount in the PV power generation equipment 110. The values ​​of charging efficiency α, discharging efficiency β, and self-discharge amount γ may be stored in the storage unit 1080 and read from the storage unit 1080 when the reduction unit 1050 performs the reduction process.

[0105] Returning to Figure 3, the determination unit 1060 determines whether the series of weight reduction processes performed by the weight reduction unit 1050 has been completed.

[0106] The charge / discharge plan determination unit 1070 determines the charge / discharge plan update value to be the charge / discharge plan that maximizes the revenue from selling electricity by discharging the battery 120, according to the determination result when the determination unit 1060 determines that the series of reduction processes by the reduction unit 1050 has been completed, and stores it in the ninth memory area 1089 of the memory unit 1080.

[0107] The output unit 1090 outputs the charge / discharge plan update value read from the ninth memory area 1089 to the battery control device 130 via the communication network 180.

[0108] Figure 9 is a block diagram showing an example of the hardware of an information processing device 10 that realizes the functions of the charge / discharge planning device 100 shown in Figure 3. The information processing device 10 comprises a processor 11, main memory 12, auxiliary memory 13, input device 14, output device 15, and communication device 16. The information processing device 10 is, for example, a personal computer, office computer, various server devices, a general-purpose machine, etc. The information processing device 10 may be implemented in whole or in part using virtual information processing resources provided using virtualization technology, such as a virtual server provided by a cloud system. The charge / discharge planning device 100 may be implemented using multiple information processing devices 10 that are connected to each other in a communicative manner.

[0109] The processor 11 is composed of, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a GPU (Graphics Processing Unit), an FPGA (Field Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), an AI (Artificial Intelligence) chip, and the like.

[0110] The main memory 12 is a device for storing programs and data, and is, for example, ROM (Read Only Memory), RAM (Random Access Memory), or non-volatile memory (NVRAM (Non-Volatile RAM)).

[0111] The auxiliary storage device 13 is, for example, an SSD (Solid State Drive), a hard disk drive, an optical storage device (CD (Compact Disc), DVD (Digital Versatile Disc), etc.), a storage system, an IC card, an SD card, a reader / writer for optical recording media, or the storage area of ​​a cloud server. Programs and data can be read into the auxiliary storage device 13 via a recording media reader or a communication device 16. Programs and data stored in the auxiliary storage device 13 are read into the main memory 12 as needed.

[0112] The input device 14 is an interface that accepts input from an external source, and can be, for example, a keyboard, mouse, touch panel, card reader, pen-input tablet, or voice input device.

[0113] The output device 15 is an interface that outputs various information such as processing progress and processing results. The output device 15 may be, for example, a display device that visualizes the above information (LCD (Liquid Crystal Display), graphics card, etc.), a device that converts the above information into sound (speaker, etc.), or a device that converts the above information into text (printer, etc.). The information processing device 10 may also be configured to input and output information to and from other devices via the communication device 16.

[0114] The input device 14 and the output device 15 constitute a user interface for receiving and presenting information with the user.

[0115] The communication device 16 is a device that enables communication (wired or wireless communication) with other devices via a communication infrastructure such as the communication network 5, and is configured using, for example, a NIC (Network Interface Card), a wireless communication module, a USB module, etc.

[0116] Furthermore, the information processing device 10 may have, for example, an operating system, a file system, a DBMS (Database Management System) (relational database, NoSQL, etc.), a KVS (Key-Value Store), etc. installed on it.

[0117] The functions of the charge / discharge planning device 100 are realized either by the processor 11 of the information processing device 10 reading and executing a control program stored in the main memory 12, or by the functions of the hardware (FPGA, ASIC, AI chip, etc.) that constitutes the charge / discharge planning device 100 itself. For example, the input unit 1010 is realized by the input device 14, the output unit 1090 is realized by the output device 15, the power price correction unit 1020, the power generation correction unit 1030, the charge / discharge planning unit 1040, the reduction unit 1050, the determination unit 1060, and the charge / discharge planning determination unit 1070 are realized by the processor 11, and the storage unit 1080 is realized by the main memory 12 and the auxiliary storage device 13.

[0118] Figure 10 is a flowchart showing an example of the process when the charge / discharge plan creation device 100 creates a charge / discharge plan for the battery 120 on the day for which the charge / discharge plan is to be created.

[0119] First, the charge / discharge plan creation device 100 determines whether a predicted value for electricity price has been input to the input unit 1010 from the electricity price prediction device 160, and whether a predicted value for the power generated by the PV power generation equipment 110 has been input to the input unit 1010 from the power generation power prediction device 170, for the days for which the charge / discharge plan for the storage battery 120 is to be created (S4010).

[0120] If no predicted values ​​for electricity prices and predicted values ​​for the power generated by the PV power generation equipment are entered into the input unit 1010 (S4010: NO), the charge / discharge plan creation device 100 repeatedly executes the process of step S4010 described above.

[0121] When the predicted value of the electricity price and the predicted value of the power generated by the PV power generation equipment are input to the input unit 1010 (S4010: YES), the electricity price correction unit 1020 instructs the first learning model 1082A to correct the uncorrected predicted value of the electricity price so that the revenue from selling electricity when the storage battery 120 is discharged is maximized. The first learning model 1082A corrects the uncorrected predicted value of the electricity price in response to the instruction from the electricity price correction unit 1020 (S4020). In addition, the power generation correction unit 1030 instructs the second learning model 1084A to correct the uncorrected predicted value of the power generation in so that the revenue from selling electricity when the storage battery 120 is discharged is maximized. The second learning model 1084A corrects the uncorrected predicted value of the power generation in response to the instruction from the power generation correction unit 1030 (S4030).

[0122] Next, the charge / discharge plan creation unit 1040 uses the corrected predicted electricity price and the corrected predicted power generation to create the initial charge / discharge plan values ​​for the target date, as shown in Figure 2 (S4040).

[0123] Next, in the reduction unit 1050, as shown in Figures 8A to 8H, a reduction process is performed in order of the discharge frames with the lowest predicted power prices after correction, taking into account the charging efficiency α, discharge efficiency β, and self-discharge amount γ, so that the planned value of the discharge amount before correction for discharge frames D1 to D7 during the discharge period becomes the dischargeable amount (S4050), and an updated charge / discharge plan value is created (S4060).

[0124] The output unit 1090 outputs this charge / discharge plan update value to the battery control device 130 via the communication network 180 (S4070). As a result, the battery control device 130 controls the charging and discharging of the battery 120 so that power bidding is carried out based on the charge / discharge plan update value.

[0125] As described above, the charge and discharge plan creation device 100 creates a charge and discharge plan for a storage battery 120 that charges the storage battery 120 with electricity generated by a PV power generation facility 110 that generates electricity using solar power as one of the renewable energy sources, and discharges the charged electricity. The device includes a charge and discharge plan creation unit 1040 that calculates the amount of charge in multiple charging units that charge the storage battery 120 and the amount of discharge in multiple discharge units that discharge from the storage battery 120, based on the predicted value of electricity prices and the predicted value of electricity generated by the PV power generation facility 110 in each time slot, which is composed of predetermined time intervals of a day, and a reduction unit 1050 that reduces the amount of discharge in each discharge unit so that the sum of the discharge amounts in the multiple discharge units becomes the amount of dischargeable, which is calculated based on the total electricity generated by the PV power generation facility 110 and the efficiency of the storage battery 120.

[0126] The charge / discharge planning device 100 makes it possible to create a highly accurate charge / discharge plan for the battery 120 that maximizes revenue from selling electricity.

[0127] Furthermore, in the charge / discharge plan creation device 100, the reduction unit 1050 calculates the dischargeable amount of the battery 120 based on the total power generated by the PV power generation equipment 110, the charging efficiency of the battery 120, the discharge efficiency, and the self-discharge amount of the battery 120.

[0128] According to the charge / discharge plan creation device 100, the charging efficiency, discharging efficiency, and self-discharge amount of the storage battery 120 are included as factors when calculating the dischargeable amount, making it possible to create a highly accurate charge / discharge plan for the storage battery 120 that maximizes revenue from selling electricity.

[0129] Furthermore, in the charge / discharge plan creation device 100, the reduction unit 1050 prioritizes the discharge units with lower predicted electricity prices and reduces the discharge amount of each discharge unit.

[0130] According to the charge / discharge planning device 100, the discharge amount is reduced in order of the discharge time slots with the lowest predicted electricity prices, making it possible to create a highly accurate charge / discharge plan for the battery 120 that maximizes revenue from selling electricity.

[0131] Furthermore, in the charge / discharge plan creation device 100, the charging time slots are time slots that fall within a period when the predicted value of the electricity price on the day for which the charge / discharge plan is created is less than or equal to the first price, and the discharge time slots are time slots that fall within a period when the predicted value of the electricity price on the day for which the plan is created is greater than or equal to the second price, which is greater than or equal to the first price.

[0132] The charge-discharge plan created by the charge-discharge plan creation device 100 is a plan in which, on the day for which the charge-discharge plan for the storage battery 120 is created, the storage battery 120 is charged all at once during the daytime when the predicted value of electricity prices is relatively low, and the storage battery 120 is discharged all at once during the nighttime when the predicted value of electricity prices is relatively high. As a result, the storage battery 120 does not have to repeat the charge-discharge operation frequently, and it is possible to prevent the premature decline in charge efficiency and discharge efficiency.

[0133] Furthermore, in the charge / discharge plan creation device 100, the reduction unit 1050 includes as one of its functions a first correction unit that corrects the values ​​of charging efficiency, discharge efficiency, and self-discharge amount based on the difference between the sum of the actual discharge amounts of multiple discharge frames on the day for which the charge / discharge plan is created and the available discharge amount.

[0134] According to the charge / discharge planning device 100, the charging efficiency, discharging efficiency, and self-discharge amount are corrected according to the aging degradation status of the storage battery 120, making it possible to create a highly accurate charge / discharge plan for the storage battery 120 that maximizes revenue from selling electricity.

[0135] Furthermore, the charge / discharge plan creation device 100 includes an electricity price correction unit 1020 that corrects the predicted electricity price on the day the charge / discharge plan is created based on the relationship between predicted and actual electricity prices in the past, and a power generation correction unit 1030 that corrects the predicted power generation of the PV power generation facility 110 on the day the charge / discharge plan is created based on the relationship between predicted and actual power generation of the power generation facility in the past. The charge / discharge plan creation unit 1040 calculates the charge amount and discharge amount for each of the multiple charging frames and multiple discharge frames based on the corrected predicted electricity price on the day the charge / discharge plan is created and the corrected predicted power generation of the PV power generation facility 110 on the day the charge / discharge plan is created.

[0136] According to the charge / discharge plan creation device 100, the predicted value of the electricity price on the target date is corrected according to the past prediction trend of electricity prices, and the predicted value of the power generation capacity of the power generation capacity on the target date is corrected according to the past prediction trend of power generation capacity of the power generation capacity. This makes it possible to create a highly accurate charge / discharge plan for the storage battery 120 that maximizes revenue from selling electricity.

[0137] Furthermore, in the charge / discharge plan creation device 100, the power price correction unit 1020 corrects the predicted power price for the day for which the charge / discharge plan is created, based on the first learning model 1082A generated by learning the relationship between predicted and actual power prices in the past.

[0138] According to the charge / discharge plan creation device 100, the predicted value of the electricity price on the target day is corrected using the first learning model 1082A, making it possible to create a highly accurate charge / discharge plan for the storage battery 120 that maximizes revenue from selling electricity.

[0139] Furthermore, in the charge / discharge plan creation device 100, the power generation correction unit 1030 corrects the predicted power generation value of the PV power generation facility 110 on the day for which the charge / discharge plan is created, based on a second learning model 1084A generated by learning the relationship between the predicted and actual power generation values ​​of the PV power generation facility 110 in the past.

[0140] According to the charge / discharge plan creation device 100, the second learning model 1084A is used to correct the predicted power generation value of the PV power generation facility 110 on the target day, making it possible to create a highly accurate charge / discharge plan for the storage battery 120 that maximizes revenue from selling electricity.

[0141] This embodiment is provided to facilitate understanding of the present invention and is not intended to limit its interpretation. The present invention may be modified or improved without departing from its spirit, and equivalents thereof are also included. [Explanation of Symbols]

[0142] 1. Power generation system 10 Information Processing Devices 11 processors 12 Main storage 13 Auxiliary storage device 14 Input devices 15 Output device 16. Communication equipment 100 Charge / Discharge Planning Device 110 PV power generation facilities 120 Battery Storage 130 Battery Control Device 150 Weather Information Provisioning Device 160 Electricity Price Prediction Device 170 Power Generation Prediction Device 180 Communication Networks 190 Power system 1010 Input Section 1020 Electricity Price Adjustment Department 1030 Power Generation Correction Unit 1040 Charge / Discharge Planning Department 1050 Weight loss section 1060 Judgment section 1070 Charge / Discharge Plan Determination Unit 1080 storage section 1081~1089 1st storage area ~ 9th storage area 1090 Output section

Claims

1. A charge / discharge planning device that creates a charge / discharge plan for a battery that charges the electricity generated by a power generation facility that uses renewable energy and discharges the charged electricity, A charge / discharge planning unit calculates the amount of charge in multiple charging periods for charging the battery and the amount of discharge in multiple discharge periods for discharging the battery, based on the predicted value of the electricity price in each period, which is formed by dividing the day into predetermined time intervals, and the predicted value of the power generated by the power generation equipment. A reduction unit reduces the difference between the total discharge amount in the multiple discharge slots and the dischargeable amount, prioritizing the discharge slot with the lower predicted electricity price among the multiple discharge slots until the total discharge amount in the multiple discharge slots reaches the dischargeable amount, which is the corrected planned value of the discharge amount, taking into account the charging efficiency, discharge efficiency, and self-discharge amount of the storage battery. A charge / discharge planning device that includes this feature.

2. A charge / discharge plan creation device according to claim 1, The charging time slot is a time slot that falls within a period when the predicted value of the electricity price on the day for which the charge / discharge plan is created is less than or equal to the first price. The discharge time slot is one that falls within a time period in which the predicted electricity price on the target date is equal to or greater than the second price, which is equal to or greater than the first price. Charge / discharge planning device.

3. A charge / discharge plan creation device according to Claim 1, The system further includes a first correction unit that corrects the values ​​of the charging efficiency, the discharging efficiency, and the self-discharge amount based on the difference between the sum of the actual discharge amounts of the multiple discharge pieces on the day for which the charge-discharge plan is created and the available discharge amount. Charge / discharge planning device.

4. A charge / discharge plan creation device according to claim 1, A second correction unit corrects the predicted value of the electricity price on the day for which the charge / discharge plan is created, based on the relationship between predicted and actual electricity prices in the past. A third correction unit corrects the predicted value of the power generation of the power generation facility on the day for which the charge / discharge plan is created, based on the relationship between the predicted and actual values ​​of the power generation of the power generation facility in the past, It further includes, The charge / discharge plan creation unit calculates the charge amount and the discharge amount based on the corrected predicted electricity price for the target date and the corrected predicted power generation of the power generation equipment for the target date of the charge / discharge plan creation. Charge / discharge planning device.

5. A charge / discharge plan creation device according to Claim 4, The second correction unit learns the relationship between predicted and actual electricity prices in the past. Based on the first learning model generated, the predicted value of the electricity price on the day for which the charge / discharge plan is created is corrected. Charge / discharge planning device.

6. A charge / discharge plan creation device according to claim 4, The third correction unit corrects the predicted value of the power generation of the power generation facility on the day for which the charge / discharge plan is created, based on a second learning model generated by learning the relationship between the predicted and actual values ​​of the power generation of the power generation facility in the past. Charge / discharge planning device.

7. A method for creating a charge and discharge plan for a battery that charges with electricity generated by a power generation facility that uses renewable energy, and discharges the charged electricity, Based on the predicted electricity price and the predicted power generated by the power generation equipment in each time slot, which is formed by dividing the day into predetermined time intervals, the amount of charge in multiple charging time slots for charging the battery and the amount of discharge in multiple discharge time slots for discharging from the battery are calculated. The difference between the total discharge amount in the multiple discharge slots and the dischargeable amount is reduced by prioritizing the discharge slot with the lowest predicted electricity price among the multiple discharge slots until the total discharge amount in the multiple discharge slots becomes the planned discharge amount, which is the corrected discharge amount considering the charging efficiency, discharge efficiency, and self-discharge amount of the storage battery. Method for creating a charge / discharge plan.

8. A method for creating a charge / discharge plan according to claim 7, The charging time slot is a time slot that falls within a period when the predicted value of the electricity price on the day for which the charge / discharge plan is created is less than or equal to the first price. The discharge time slot is one that falls within a time period in which the predicted electricity price on the target date is equal to or greater than the second price, which is equal to or greater than the first price. Method for creating a charge / discharge plan.

9. A method for creating a charge / discharge plan according to Claim 7, Based on the difference between the sum of the actual discharge amounts of the multiple discharge units on the day for which the charge-discharge plan is created and the available discharge amount, the values ​​of the charging efficiency, the discharge efficiency, and the self-discharge amount are corrected. Method for creating a charge / discharge plan.

10. A method for creating a charge / discharge plan according to Claim 7, Based on the relationship between past predicted and actual electricity prices, the predicted electricity price for the day on which the charge / discharge plan is created is corrected. Based on the relationship between the predicted and actual power generation values ​​of the aforementioned power generation facility in the past, the predicted power generation value of the aforementioned power generation facility on the specified date is corrected. The charge amount and discharge amount are calculated based on the predicted electricity price on the day the corrected charge / discharge plan is created and the predicted power generated by the power generation facility on the day the corrected charge / discharge plan is created. Method for creating a charge / discharge plan.

11. A method for creating a charge / discharge plan according to claim 10, Based on a first learning model generated by learning the relationship between predicted and actual electricity prices in the past, the predicted electricity price for the day for which the charge / discharge plan is created is corrected. Method for creating a charge / discharge plan.

12. A method for creating a charge / discharge plan according to claim 10, Based on a second learning model generated by learning the relationship between predicted and actual power generation values ​​of the power generation facility in the past, the predicted power generation value of the power generation facility on the day for which the charge / discharge plan is created is corrected. Method for creating a charge / discharge plan.

Images