Renewable energy operation optimization system

Abstract

To economically optimize the energy efficiency of factories and other facilities of individuals and small and medium-sized enterprises. [Solution] A server 12 and a computer 28 for controlling the EMS 30 installed in the solar power generation equipment 32 are connected by a network 26. The server 12 collects weather information 20 and predicts the amount of power generated. The server 12 delivers control data 24 for the EMS 30, which is generated individually according to the user's request, to the user's computer 28. The control data 24 includes data for setting the start and end times for selling electricity, buying electricity, and storing electricity, and setting data 38 for selecting the operating mode of electrical equipment 34 that consumes the generated power. The server 12 delivers the control data 24 based on the latest power generation forecast calculation at short update intervals.

Description

Technical Field

[0001] The present invention relates to a system that integrally controls a photovoltaic power generation facility, a power storage facility, a load facility, and power trading in order to optimize the operation of natural energy.

Background Art

[0002] In power generation facilities equipped with a photovoltaic power generation device and a storage battery, in order to send the generated power to a predetermined supply destination, a power management device (EMS, energy management system) equipped with a power conditioner and its operation control device is used (Patent Document 1). Further, a system has been developed in which a computer of a power management device for a home or facility equipped with such a photovoltaic power generation device and a storage battery is connected to a server computer of an operator via a computer network, and power supply and demand monitoring and data analysis are performed on the server side (Patent Document 2).

[0003] Furthermore, a system has been developed in which a server and a home HEMS (home energy management system) controller in a house are connected via a network to enable remote operation of electrical appliances in the house (Patent Document 3). Also, a system has been developed to optimize the power reception management by electrical equipment and storage batteries of power consumers (Patent Document 4). In addition, a technique for predicting power demand has also been developed (Patent Document 5).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Patent Document 5

[0005] In typical household solar power generation systems, the generated electricity is consumed within the home, and any surplus is sold back to the grid. However, with the proliferation of large-scale solar power generation facilities, there are cases where there is a surplus of electricity during the day, making it impossible to sell it back to the grid. Therefore, measures such as storing surplus electricity in batteries and controlling the system while considering the selling price and buying price of electricity are being implemented. However, as pointed out in the aforementioned patent document, the factors that fluctuate are diverse, making it difficult to optimize the management and operation of the system.

[0006] This invention was made in view of the above points, and aims to provide a system that accurately analyzes a wide variety of factors on a server with high computing power and delivers control data that can optimize the balance of electricity sales and storage / discharge for solar power generation equipment in private homes and large-scale renewable energy power generation systems. [Means for solving the problem]

[0007] The following configurations are each means of solving the above-mentioned problems.

[0008] <Configuration 1> The server 12 and the computer 28, which controls the EMS 30 (energy management system) installed in the user-side power generation equipment 32 that utilizes natural energy, are connected by a network 26. Server 12 collects weather information 20 and facility environment information 22, performs environmental prediction calculations necessary for controlling the EMS 30, and distributes control data 24 for controlling the EMS 30 to the user's computer 28. The control data 24 described above includes data for setting the start and end times for selling, buying, storing, and discharging electricity, and setting data 38 for setting the control of electrical equipment 34 that consumes power. The user-side EMS30 receives the control data24 from the computer28 and controls the electrical equipment34 connected to the power generation equipment32. A renewable energy operation optimization system characterized in that the server 12 sets the maximum acceptable delay time from when it distributes control data 24 to the computer 28 mentioned above until it performs a new environmental prediction calculation and distributes new control data 24 as the data update time, and distributes the latest control data 24 based on the latest environmental prediction calculation to the user's computer 28 at each update time.

[0009] <Configuration 2> The renewable energy operation optimization system according to Configuration 1, characterized in that the data update time is 1 hour or less.

[0010] <Structure 3> The renewable energy operation optimization system according to Configuration 1, characterized in that the control data 24 described above includes the period operation schedule of electrical equipment 34 connected to the power generation equipment 32 described above for a predetermined period, and this control data 24 is updated and distributed at the above update time.

[0011] <Structure 4> The renewable energy operation optimization system according to Configuration 1, characterized in that the user's computer 28 uses the latest control data when it receives control data 24 from the server 12, and if it fails to receive control data, it uses the control data 24 received immediately before to have the EMS 30 continue control.

[0012] <Composition 5> The renewable energy operation optimization system according to Configuration 1, characterized in that the server 12 receives a designation from the user's computer 28 to prioritize either optimizing the balance of electricity sales and purchases or optimizing CO2 emissions, and distributes the corresponding control data 24 to the user's computer 28.

[0013] <Composition 6> The server 12 is configured to generate and distribute control data 24 for each user to control the state of the specified electrical equipment 34 on the user side, as described in Configuration 1 of the natural energy operation optimization system.

[0014] <Configuration 7> The server 12 distributes control data for controlling the EMS 30 of a plurality of facilities specified by the user side, generates and distributes control data that enables optimization of the electricity trading balance or optimization of the CO2 emission amount when the above-mentioned plurality of facilities are combined, as described in Configuration 1 of the natural energy operation optimization system.

[0015] <Configuration 8> The server 12 acquires data indicating the states of the power generation equipment 32 and the electrical equipment 34 on the user side at the same timing as the measurement and uses it for environmental prediction calculation, as described in Configuration 1 of the natural energy operation optimization system.

[0016] <Configuration 9> A method for predicting future power generation amounts, characterized in that a theoretical power generation value (future) is calculated based on weather forecast data, the time series data thereof is predicted by a machine learning model, and a corrected power generation prediction value is obtained. On the other hand, the next power generation amount is predicted by a regression type machine learning model using the theoretical power generation value (actual) calculated from the weather performance data or the power generation performance data as an input.

[0017] <Configuration 10> The natural energy operation optimization system according to any one of Configurations 1 to 8, characterized in that control data is obtained using the future power generation amount obtained by the method according to Configuration 8 or 9.

Advantages of the Invention

[0018] <Advantages of Configuration 1> Since the server can execute high-precision environmental prediction calculations frequently based on the latest valid information and use the results to control the equipment on the user side, the energy usage efficiency of individuals, factories of small and medium-sized enterprises, etc. can be economically optimized. <Effect of Configuration 2> If control data is distributed at short time intervals of 1 hour or less, for example, 30 minutes or 15 minutes, optimal control at the present moment can be achieved by utilizing the high computing power of the server. <Effect of Configuration 3> Since control data is received for a certain period and its content is updated in a short time, optimal operation can always be achieved. At the same time, even if a communication failure occurs, the operation will not be interrupted. <Effect of Configuration 4> Since the control data is updated in a short time, stable control can be achieved with little variation even when using the control data received immediately before. <Effect of Configuration 5> Control data can be distributed to meet the operation purposes desired by each user. <Effect of Configuration 6> When the user specifies the electrical equipment to be controlled by EMS30, individual distribution of optimal control data according to the user's circumstances can be achieved. <Effect of Configuration 7> When a user owns independent facilities at multiple locations, they can be regarded as one facility for optimization calculation, and control data for each facility can be distributed to facilitate the user. <Effect of Configuration 8> If the status of the user's power generation equipment and electrical equipment can be obtained at any time, the user's requests can be faithfully responded to.

Brief Description of Drawings

[0019] ​​​​​​​​​​​​ [Modes for carrying out the invention]

[0020] The embodiments of the present invention will be described in detail below for each example. [Examples]

[0021] (System Overview) Figure 1 is a block diagram showing an overview of the operational optimization system for the renewable energy power generation facility 32 in Example 1. As shown in Figure 1, the electricity generated by the solar power generation equipment 32 is supplied to electrical equipment 34, and any surplus electricity is used to charge the storage battery 31 or to sell the electricity back to the grid. The EMS 30 (Energy Management System) is a device for controlling the distribution of this electricity, and the computer 28 is provided to optimize the operation of the EMS 30. The computer 28 and the EMS 30 may be separate or integrated units. The computer 28 is connected to the server 12 via the network 26.

[0022] (server) Server 12 collects weather information 20, load power 17, and solar power generation 19, performs predictive calculations necessary for controlling the EMS 30, and distributes control data 24 for controlling the EMS 30 to the user's computer 28. For this purpose, as shown in Figure 1, Server 12 is equipped with an environmental prediction calculation unit 14, a communication unit 16, an information collection unit 18, an optimization control calculation unit 23, etc. in its calculation processing unit. In this invention, Server 12 reads user-specific data 54 and provides a service that generates and distributes control data 24 optimized for each user. Load power 17 is the amount of electricity consumed by the user's electrical equipment at regular intervals. Solar power generation 19 is the amount of electricity generated by the solar power generation equipment at regular intervals.

[0023] The memory unit of server 12 stores weather information 20 and equipment environment information 22 collected by the information collection unit 18. Weather information 20 includes information necessary for calculating power generation, such as past weather, temperature, and sunshine duration. Equipment environment information includes data on the operating history of the user's equipment and measurement data of power consumption. The operating history of the equipment includes data on when the equipment started and stopped and what settings it was operated under. Environmental forecasts include forecasts of solar power generation amount 19 and load power amount 17. Weather information 20 and equipment environment information 22 are used in the calculation processing of the environmental forecast calculation unit 14.

[0024] (User side) Furthermore, as shown in Figure 1, the user's request items are stored in the storage device of the user's computer 28. These are read by the communication unit 16 of the server 12 and stored in the server 12's storage as user-specific data 54. The optimization control calculation unit 23 generates user-specific control data 24 while referring to this user-specific data 54.

[0025] (Control data) As will be explained in more detail later, the control data 24 includes, for example, data that controls the start and end times of electricity sales, electricity purchases, and energy storage, as well as data that controls the operating mode of electrical equipment 34 that consumes generated power. The user's EMS 30 receives the control data 24 from the server 12 via the computer 28 and controls the battery 31 and electrical equipment 34 connected to the power generation equipment 32.

[0026] (Update of prediction calculation) Server 12 distributes control data 24 to the user's computer 28, then performs new environmental prediction calculations and distributes new control data 24. The maximum acceptable delay time until the new data is distributed is set as the data update time. An update interval of one day is too long due to the large changes in the natural environment during that time. If prediction calculations are performed when the sun is shining and control data is distributed after it becomes cloudy, the control will malfunction.

[0027] Therefore, the time granularity of the collected data is set to the minimum time interval for recalculation. This is set to the maximum acceptable delay time for data updates. In practice, it is preferable to operate at intervals of, for example, 30 minutes, and to further shorten the interval as the calculation speed increases. Server 12 delivers control data 24 based on the latest environmental prediction calculation to the user's computer 28 at this update time interval.

[0028] (Data specified for each user) Figure 2 is an explanatory diagram showing the relationship between user-specific data 54 and control data 24. In this invention, as described above, control data 24 tailored to the equipment and environment owned by the user is distributed from the server 12 in a specialized manner for each user. That is, the server 12 individually generates and distributes control data 24 that is optimized to the greatest extent possible by incorporating the wishes of each user. For this purpose, the storage unit of the server 12 individually stores user-specific data 54, which lists the specified contents for each user.

[0029] Here, using Figure 2, we will explain the contents of user-specific data 54 and a concrete example of the control data generated according to that specification. First, the time selection specification data 36 is data that indicates a user's preference, for example, to receive control data only between 7:00 and 18:00, due to the user's circumstances such as working hours.

[0030] Furthermore, in the specified period operation schedule 40, it is possible to specify, for example, that the distribution of control data 24 is not required during the summer vacation period, as in the case of schools. It is also possible to include requests for operation schedules such as not using the air conditioning equipment on holidays. For example, it is also possible to specify whether to use the control data 24 from the server 12 or to control it internally for a certain period.

[0031] The setting data 38 includes mode selection specification data that indicates a preference, such as wanting to control the air conditioner in only cooling / heating mode or only fan mode. If multiple air conditioners are installed, it is possible to specify the switchable operating mode for each air conditioner.

[0032] The user can identify the electrical equipment 34 to be controlled using the controlled equipment specification data 42. When various electrical equipment is installed, controlling only the equipment with the highest power consumption can be done to avoid complex control and enable efficient operation. The optimization priority item specification data 44 is data that indicates, for example, whether to prioritize minimizing CO2 emissions or maximizing electricity sales profits 52.

[0033] In addition to the above, in this embodiment, the user's computer 28 notifies the server of equipment environment notification data 46 and operational status notification data 48. The equipment environment notification data 46 contains information about the contents and capacity of the renewable energy power generation equipment owned by the user. There are various types of power generation equipment, such as solar power generation equipment, wind power generation equipment, and geothermal power generation equipment, but it is good to notify the server with data that shows what level of power generation capacity is available under what specific conditions.

[0034] For example, it would be good to notify users of their power generation performance over a certain period. It would also be good to notify them of the electricity purchase price from the power company they are contracted with, the electrical equipment they own, the maximum possible power generation, and the battery capacity. The operational status notification data 48 should include the usage history of the user's electrical equipment. In addition, it can include data to update changes in the equipment environment that have been reported. For example, it can include changes in the sunlight conditions at the location where the user's electrical equipment is installed.

[0035] The server receives actual data on power demand and power generation. This is to optimize the forecast calculations. CO2 emission data 50 is data calculated from CO2 emissions at regular intervals. Electricity sales profit data 52 is data calculated from electricity sales data obtained from the meter. The details of the electricity supply contract between the user and the power company are also notified. If the formula for calculating electricity sales profit 52 is known, it would be good to notify this as well.

[0036] (Control data) In response to the user's specifications as described above, control data as shown in Figure 2 is distributed. First, control data is distributed to control the start and end times of electricity sales, electricity purchases, and battery charging and discharging. In accordance with the setting data 38 for temperature settings and mode selection, control data is distributed to control the set temperature of each air conditioner that consumes the generated power, the time to switch between heating / cooling mode and fan mode, and the operating schedule.

[0037] Furthermore, the timing of the distribution of control data 24 is controlled in accordance with the specified period operation schedule 40, so as to be linked to the calendar. In addition, data is distributed to control the on / off timing, etc., of each electrical device specified by the user, in accordance with the specified control target equipment data 42.

[0038] It can generate control data for electrical equipment such as ventilation fans and blower fans in packaged air conditioners (EHP: electric heat pumps), large-scale air conditioners (AHU: Air Handling Units), and air conditioning pumps (hot and cold water pumps) used in buildings and offices. It can also generate time connection data and mode control data used for controlling the stopping time (control rate) of outdoor units, etc., for dedicated devices that control the intermittent operation of air conditioning outdoor units.

[0039] For example, it includes a predetermined operating schedule for electrical equipment 34 connected to the power generation equipment 32 over a set period of several days. This control data 24 is updated and distributed at the above-mentioned update times. In addition, the user's computer 28 always uses the latest control data 24 when it receives it from the server 12, and if it fails to receive it, it can use the control data 24 that was received immediately before to allow the EMS 30 to continue control.

[0040] When optimization priority item designation data 44 is received, indicating that either the optimization of the balance of electricity sales or the optimization of CO2 emissions should be prioritized, the system calculates, for example, CO2 emissions and electricity sales profits 52 and distributes control data to optimize either of them along with the results.

[0041] Furthermore, users may own multiple facilities in different locations. In this case, control data can be generated and distributed independently for each facility, but it is also possible to optimize control by comprehensively managing the equipment of multiple facilities. Such control cannot be achieved with individual control by EMS30 installed at each facility. Servers 1 and 2 can generate and distribute control data that optimizes the balance of electricity sales and purchases or CO2 emissions when multiple facilities specified by the user are combined. Control data can also be generated for each building or contract unit.

[0042] (Server calculation process) Figure 3 is an explanatory diagram of the process on the server that performs predictive calculations based on the conditions requested by the user and generates control data. The central function f(t) represents the sum of each energy quantity at a given time t, and is 0 at that time t. f(t) = Predicted power generation (t) - Predicted demand (t) - Planned charging (t) + Planned discharging (t) + Planned energy saving (t) + Planned electricity purchase (t) - Planned electricity sales (t) = 0 Each variable is labeled with a positive or negative sign, where positive is the amount of electricity input (or obtained) for the facility in question. Charging is output to the battery side, so it is negative. Electricity sold is also output to the power company (grid side), so it is negative. In the function f(t), the environmental forecast data is determined first, and the server's environmental forecast calculation unit 14 (Figure 1) performs detailed forecast calculations using parameters such as forecasts of solar power generation 19 considering climate change, forecasts of electricity demand considering the season, and trading prices from the Japan Electric Power Exchange.

[0043] The objective functions for minimizing electricity costs and CO2 emissions are created using a function f(t) at a certain time t, and a simulation is performed to minimize the sum or integral value over a specified period (t1~t2) according to the respective conditions. The variables are set as follows: • Input values: Predicted power generation, predicted demand • Planned electricity purchase amount • Planned electricity sales volume • Planned energy savings (Constraints: Control conditions for load equipment) • Planned charge amount (constraints: battery specifications) • Planned discharge amount (constraints: battery specifications) Objective function for minimizing electricity costs = Σt1→t2 {Planned electricity purchase amount (t) × electricity purchase price - Planned electricity sales amount (t) × electricity sales price} While the unit prices are listed in a simplified form, it is possible to include variable elements such as fuel cost adjustments and renewable energy surcharges. Furthermore, it is possible to consider a basic charge based on maximum demand power to match the facility's electricity contract. Objective function for minimizing CO2 emissions (maximizing self-consumption) = Σt1→t2 {Planned electricity purchase amount (t)} = If you want to maximize self-consumption, and only store generated electricity, you can calculate self-consumption from the amount of electricity generated and the amount of battery discharged. However, if you consider the condition of charging purchased electricity, the total amount of planned purchased electricity over the period must be minimized.

[0044] The control plan is determined by simulating each planned quantity, such as planned electricity purchase, planned electricity sales, planned energy savings, planned charging, and planned discharge, so as to minimize the objective function, "electricity cost" or "CO2 emissions." For the simulation, mathematical optimization methods such as particle swarm optimization and genetic algorithms can be employed.

[0045] The environmental forecast calculation unit 14 collects weather information and calculates electricity demand forecasts that take the season into account, as well as standard predicted power generation amounts. Furthermore, it predicts the amount of solar power generation 19 for the relevant user based on a comparison of these predicted values ​​with the user's actual power generation results. At this time, it obtains the user's equipment usage history, battery remaining capacity, and operating plan from the equipment environment notification data 46 and the operational status notification data 48 and uses them in the forecast calculation.

[0046] (Power generation forecast calculation) The above process requires the calculation of predicted power generation with the highest possible accuracy. Figure 4 illustrates this calculation procedure. Weather forecast data, including calendar information, sunshine duration, and temperature, is used for the prediction calculation.

[0047] (Electricity generation forecast calculation) In the calculation shown in Figure 4, the right-hand section calculates theoretical power generation values ​​based on weather forecast data. For example, this calculation is performed every 30 minutes to obtain time-series data of theoretical power generation values ​​(future). Then, a machine learning model (deep neural network) is used to correct the theoretical power generation values ​​and obtain predicted power generation values. This machine learning model is pre-trained with theoretical power generation values ​​calculated from actual weather data as explanatory variables and measured actual power generation values ​​as the dependent variable. This involves predicting the amount of power generation, which is also the dependent variable in the future, from the theoretical power generation values ​​of the future as explanatory variables, and is a correction process based on the universal approximation theorem of neural networks. As a result, it is possible to obtain predicted power generation values ​​that take into account the characteristics and failure conditions of power generation equipment, which cannot be obtained from weather forecast data alone.

[0048] On the other hand, the left side of Figure 4 acquires time-series data of actual power generation. If actual power generation data is not acquired, the theoretical power generation value is calculated based on actual weather data. Here, the time-series data of the theoretical power generation value (actual) is obtained. Here as well, the same machine learning model (deep neural network) as above is used to correct the calculation result of the theoretical power generation value. In this way, the corrected theoretical power generation value (actual) is obtained.

[0049] The above process yields time-series data of theoretical (actual) power generation values. By using a recurrent machine learning model (recurrent neural network), recurrent prediction becomes possible, allowing for power generation to be predicted using methods other than those described above. If actual power generation data exists, the machine learning model will input both actual power generation data and weather data. However, if actual power generation data does not exist, only the corrected theoretical (actual) power generation values ​​will be input. This is because the latter includes elements of weather data. [Examples]

[0050] Ultimately, the accuracy of power generation forecasts is improved by ensembling predictions based on future weather forecasts (deep neural network) and predictions based on past performance (recurrent neural network). Furthermore, if recent power generation data and weather data are regularly available, it becomes possible to periodically revise the forecast using these as new input data.

[0051] Patent Document 5 describes a technique for generating predictive data showing a user's future energy demand forecast in a time series, using a neural network machine learning model, from multiple user data showing the user's past energy demand performance in a time series. Known techniques like this can be used for the machine learning model. In this way, future power generation is output, for example, every 30 minutes, and used for calculating the control data shown in Figure 3.

[0052] Furthermore, load control (energy-saving control) can be set arbitrarily and included in the objective function. By comparing predictions with actual results, fault detection of power generation equipment and electrical devices on the user's side is also possible.

[0053] The actual controls involve the charging and discharging of storage batteries, electricity purchase and sale, and the control of load equipment, integrating power generation, storage, and energy saving. Optimizing the operation of power generation equipment means minimizing costs (minimizing "electricity purchase price - electricity sales profit") and maximizing the utilization rate of renewable energy. This refers to minimizing CO2 emissions. To achieve this, a "control plan" is implemented for each piece of equipment.

[0054] The data used in the above system is, for example, as follows: (Data specifying the controlled device) This data is specified for each user and includes information such as which devices will be controlled. This data includes basic specification information necessary for simulation (such as rated power consumption and set value range).

[0055] (Load equipment schedule control specification data) This is not an automatically planned control plan, but rather a control plan that the user decides on at their discretion. (Optimization priority item specification data) This data specifies whether cost or CO2 emissions should be prioritized. (Optimization range setting data) This data indicates whether an optimization plan should be created within a specified range. It specifies whether the plan should be limited to specific equipment, apply to all customers, or combine multiple customers.

[0056] (Equipment control data) This is schedule data that determines the date and time for control. For example, in the case of a battery storage system, you would determine the timing for switching between charge and discharge modes and setting upper and lower limits for charging. Furthermore, you would determine the timing for controlling electricity sales and purchases. You would also determine the timing for starting and stopping load equipment, controlling set values, starting and stopping air conditioners, mode switching, and temperature settings.

[0057] The above-described present invention enables the following optimizations. (Optimization of operations) This allows for the optimization of control plans for devices such as battery charging and discharging, electricity purchase and sale, and the start / stop and setting changes of load equipment. (Cost optimization) This allows for minimizing the "electricity purchase cost - electricity sales profit" (fee balance). (Optimization of CO2 emissions) Minimizing the amount of electricity purchased maximizes the utilization rate of renewable energy. The turnover rate of battery storage increases, and the amount of electricity sold back to the grid decreases.

[0058] According to the system described above, for example, it is possible to send customized control data to specific registered users, enabling control that is appropriate for each user's electricity purchase price from their contracted power company, owned electrical equipment, maximum possible power generation, and battery capacity.

[0059] Furthermore, by uniformly controlling a group of members in a designated area, it can be operated as a large-capacity power generation facility in that area. By controlling this large-capacity power generation facility according to its operational purpose, it is possible to optimize CO2 emissions and maximize the efficiency of renewable energy utilization overall. When a large number of users are ordinary households, different trends from inter-power company power sharing control can be observed, enabling control planning based on highly accurate demand and power generation forecasts.

[0060] Furthermore, even if communication between the control computer and the server is unstable, each can maintain its data and operate synchronously during periods when communication is possible. This allows for the construction of a control system that frequently repeats re-prediction and control replanning based on past performance. Additionally, it is possible to construct a system that allows for the free modification of the conditions and scope of optimal control, such as optimization for a single facility or optimization for multiple facilities as a whole. [Explanation of symbols]

[0061] 12 servers 14. Environmental Prediction Calculation Unit 16 Communications Department 17 Load power 18. Information Gathering Department 19 Solar power generation 20 Weather Information 22 Facility environment information 23 Optimization Control Calculation Unit 24 Control Data 26 Network 28 Computer 30 EMS 32 Power generation equipment 31 Storage Battery 34 Electrical equipment 36-hour time selection data 38 Configuration Data 40. Designated Operating Schedule for a 40-Period Period 42. Data specifying the controlled device 44. Data specifying optimization priority items 46. ​​Equipment Environment Notification Data 48. Data for operational status notification 50 CO2 emission figures 52. Profits from selling electricity 54. Data specified for each user

Claims

1. The server and the computer used to control the EMS (energy management system) installed in the user's power generation equipment that utilizes renewable energy are connected via a network. The server collects weather information and facility environment information, performs environmental prediction calculations necessary for controlling the EMS, and distributes control data to the user's computer for controlling the EMS. The control data described above includes data for setting the start and end times for selling, buying, storing, and discharging electricity, as well as setting data for controlling electrical equipment that consumes power. The user-side EMS mentioned above receives the control data from the computer and controls the electrical equipment connected to the power generation facility mentioned above. A renewable energy operation optimization system characterized by a server that sets the maximum acceptable delay time between distributing control data to the aforementioned computer and performing new environmental prediction calculations and distributing new control data as the data update time, and then distributing control data based on the latest environmental prediction calculations to the user's computer at each update time.

2. The renewable energy operation optimization system according to claim 1, characterized in that the data update time is one hour or less.

3. The renewable energy operation optimization system according to claim 1, characterized in that the control data described above includes the period operation schedule of electrical equipment connected to the power generation facility described above for a predetermined period, and this control data is updated and distributed at the above update time intervals.

4. The renewable energy operation optimization system according to claim 1, characterized in that the user's computer uses the latest control data when it receives control data from the server, and if it fails to receive data, it uses the control data that was received immediately before to allow the EMS to continue control.

5. The renewable energy operation optimization system according to claim 1, characterized in that the server receives a designation from the user's computer to prioritize either optimizing the balance of electricity sales and purchases or optimizing CO2 emissions, and distributes the corresponding control data to the user's computer.

6. The renewable energy operation optimization system according to claim 1, characterized in that the server generates and distributes control data for each user to control the state of electrical equipment specified by the user.

7. The server distributes control data to control the EMS systems of multiple facilities specified by the user. The renewable energy operation optimization system according to claim 1, characterized by generating and distributing control data that enables optimization of the balance of electricity sales and purchases or the optimization of CO2 emissions when the above-mentioned multiple facilities are combined.

8. The renewable energy operation optimization system according to claim 1, characterized in that the server acquires data indicating the status of the user's power generation equipment and electrical devices at the same time as the measurement and uses it for environmental prediction calculations.

9. A method for predicting future power generation, characterized by calculating theoretical power generation values ​​(future) based on weather forecast data, predicting that time series data using a machine learning model, and obtaining corrected power generation prediction values, while simultaneously using a regression machine learning model to predict the next power generation amount, taking theoretical power generation values ​​(actual) or actual power generation data calculated from actual weather data as input.

10. The renewable energy operation optimization system according to any one of claims 1 to 8, characterized in that control data is obtained using the future power generation amount obtained by the method described in claim 8 or 9.

Images