Operation method and device for wind and storage combined participation in electric power spot market, and storage medium

Abstract

The invention relates to an operation method and device for wind storage combined participation in an electric power spot market, and a storage medium. The method comprises the steps that an energy storage energy management system of a wind power plant obtains wind power generation active power prediction data and electricity price data; in the day-ahead market, determining a next-day energy storage application scene according to the day-ahead electricity price by considering the full-life-cycle economy of an energy storage project, and making a day-ahead market operation strategy; in the real-time market, the change trend of the electricity price is researched and judged based on the predicted real-time electricity price and the day-ahead electricity price, the energy storage charging and discharging strategy is further adjusted, and a real-time market operation strategy is formulated; and deploying the formulated market operation strategy to an energy storage energy management system of the wind power plant for operation. Compared with the prior art, the energy storage system is integrated into a new energy station production management and marketing management system, the method is suitable for engineering practice in the starting stage of the current electric power spot market, the consumption capacity and the electricity selling income of the new energy station are improved, and the competitiveness in the electric power spot market is improved.

Description

Technical Field

[0001] The present invention relates to a new energy storage control technology, and in particular to an operation method, device and storage medium for wind and energy storage combined operation to participate in the electricity spot market based on electricity price forecast. Background Art

[0002] The output of new energy is random, intermittent and volatile, which increases the balancing cost of the power grid system and weakens the competitiveness of the new energy spot market, which is not conducive to the large-scale and high-quality development of new energy. Energy storage technology can change the spatiotemporal characteristics of electric energy, quickly respond to power demands at different time scales, and smooth out the fluctuations in new energy output, thus becoming an important supporting infrastructure for the large-scale development of new energy.

[0003] In the future, as new energy stations will gradually enter the electricity spot market, energy storage can increase the controllability of new energy, increase the competitiveness of stations in the spot market, and improve the absorption capacity and electricity sales revenue for new energy stations.

[0004] At present, the research on the joint participation of new energy storage in the electricity spot market is mainly based on theoretical research in an ideal mature electricity market. For example, the schemes disclosed in Chinese patent applications CN202310311759.0, CN202310305301.4 and CN202310357914.2 use methods such as conditional risk value to optimize the bidding strategy design of the day-ahead market and the real-time market, which is suitable for the mature electricity spot market of "quoting quantity and quotation" in the future, but not for the actual engineering of the current electricity spot market in its initial stage. Summary of the invention

[0005] The purpose of the present invention is to take into account the engineering reality at the initial stage of the current electricity spot market and to provide an operation method, device and storage medium for wind and storage joint participation in the electricity spot market based on electricity price forecast with "reporting quantity but not price".

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] As a first aspect of the present invention, a method for wind power storage to jointly participate in the electricity spot market is provided, and the method steps include:

[0008] Connect the AGC, wind power generation active power forecasting system and marketing management system data on the site side to the energy storage management system of the wind farm, and obtain wind power generation active power forecasting data and electricity price data;

[0009] In the day-ahead market, the economic feasibility of the energy storage project over its entire life cycle is considered, and the application scenarios of energy storage for the next day are determined based on the day-ahead electricity price, so as to formulate the day-ahead market operation strategy.

[0010] In the real-time market, the real-time electricity price is predicted, and the trend of electricity price changes is analyzed based on the predicted real-time electricity price and the day-ahead electricity price. The energy storage charging and discharging strategy is further adjusted to formulate a real-time market operation strategy.

[0011] The formulated market operation strategy is deployed to the energy storage energy management system for operation.

[0012] As a preferred technical solution, the day-ahead market operation strategy is formulated as follows:

[0013] Calculate the average highest and lowest hourly and average electricity prices based on the day-ahead electricity prices at the set time, and further calculate the day-ahead electricity price difference

[0014] If the product of the average ratio of the day-ahead electricity price to the real-time electricity price and the calculated day-ahead electricity price difference is greater than the energy storage kilowatt-hour cost, the energy storage will participate in the spot market the next day and discharge and charge during the hours when the day-ahead electricity price is the highest or lowest on average;

[0015] If the product of the average ratio of the price difference between the day-ahead electricity price and the real-time electricity price and the calculated day-ahead electricity price difference is less than the energy storage cost per kilowatt-hour, the energy storage will track the wind power generation active power forecast curve the next day.

[0016] As a preferred technical solution, the average ratio of the price difference between the day-ahead electricity price and the real-time electricity price is calculated based on the day-ahead electricity price and the real-time electricity price of historical data, and the price difference between the average highest electricity price in h hours and the lowest electricity price in h hours is calculated, and the ratio between the two is calculated:

[0017]

[0018] In the formula, It is the price difference between the day-ahead electricity price and the average highest and lowest real-time electricity price at h hours on a long time scale; h is the energy storage duration.

[0019] As a preferred technical solution, the electricity cost The energy storage cost is calculated by leveling the cost and discharge volume over the entire life cycle of the energy storage system:

[0020]

[0021] In the formula, the total cost of energy storage over the entire life cycle is Including the initial investment and construction costs of the project Annual operation and maintenance costs Periodic component replacement costs and cost recovery at the end of the project is the total online power consumption;

[0022]

[0023]

[0024]

[0025] In the formula, is the unit battery capacity of energy storage and the PCS power cost; is the energy storage capacity and maximum charge and discharge power; is the unit battery capacity of energy storage and the maintenance cost of PCS power; r is the discount rate; Y i For the entire life cycle of energy storage; The replacement cost of batteries per unit of energy storage; Replace the year set for the energy storage battery; i is the energy storage recovery coefficient; is the energy storage discharge depth; The number of times the energy storage is charged annually; is the energy storage self-discharge rate; It is the rate of battery capacity decay in each energy storage cycle.

[0026] As a preferred technical solution, the real-time market operation strategy is formulated as follows:

[0027] When the marketing platform announces the real-time electricity prices of the last four moments, it combines them with historical data and predicts the real-time electricity prices of the next four moments. The energy storage charging and discharging strategy is adjusted based on the predicted electricity prices:

[0028] If the predicted real-time electricity price at time t is greater than the first electricity price threshold, the energy storage will be discharged for 15 minutes;

[0029] If the predicted real-time electricity price at time t is less than the second electricity price threshold, the energy storage is charged for 15 minutes.

[0030] As a preferred technical solution, if the day-ahead electricity price at time t is greater than the first electricity price threshold but the predicted real-time electricity price is less than the third electricity price threshold, the energy storage will not be discharged, and will be discharged when the real-time electricity price at a subsequent time is detected to be greater than the first electricity price threshold. However, if there are only h hours left until 0:00 the next day, the energy storage will be discharged immediately without affecting the energy storage operation strategy for the next day; the first electricity price threshold is greater than the third electricity price threshold;

[0031] If the day-ahead electricity price at time t is less than the second electricity price threshold but the predicted real-time electricity price is greater than the fourth electricity price threshold, the energy storage will not be charged, and charging will be performed when the real-time electricity price at a subsequent moment is detected to be less than the second electricity price threshold. However, if there are only h hours left in the discharge period, immediate charging will not affect the energy storage discharge; the second electricity price threshold is less than the fourth electricity price threshold.

[0032] As a preferred technical solution, the method constructs a real-time electricity price prediction model based on a CNN-LSTM deep neural network to predict the real-time electricity price;

[0033] The real-time electricity price prediction model uses the historical real-time electricity prices of the electricity spot market since its continuous operation as input, uses a convolutional neural network to extract input features, and uses the CNN extracted features as the input of the long short-term memory network for time series prediction, and outputs the real-time electricity prices at the next four moments.

[0034] As a preferred technical solution, the real-time electricity price prediction model is as follows:

[0035] The convolutional neural network part consists of 5 layers of CNN, and each CNN layer is connected to a BatchNorm2d layer and a Dropout layer; the BatchNorm2d layer normalizes the data; the Dropout layer randomly invalidates a part of the hidden layer nodes according to a certain probability in each training;

[0036] The long short-term memory network consists of two layers of LSTM. The features extracted by the convolutional neural network are used as input for time series prediction. The first layer is used as a pre-training layer, and the second layer is the Main-LSTM. The output results pass through a fully connected layer to adjust the vector format specified for the output.

[0037] Autoregressive layer, in parallel with the CNN-LSTM part.

[0038] As a second aspect of the present invention, there is provided an operating device for the joint participation of wind and energy storage in the electricity spot market, comprising a memory, a processor, and a program stored in the memory, wherein when the processor executes the program, it implements the operating method for the joint participation of wind and energy storage in the electricity spot market based on electricity price forecast as described above.

[0039] As a third aspect of the present invention, a storage medium is provided, on which a program is stored, and when the program is executed, the method for operating the wind and storage joint participation in the electricity spot market based on electricity price forecast as described above is implemented.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1) The method of wind and energy storage jointly participating in the electricity spot market based on electricity price forecast proposed in the present invention is aimed at the new energy storage on the power supply side. It considers the use of distributed energy storage technology in wind farms. Energy storage acts as a price acceptor and participates in the electricity market in the form of reporting quantity but not quotation. The energy storage system is integrated into the production management and marketing management system of new energy stations, which is more suitable for the actual engineering in the initial stage of the current electricity spot market.

[0042] 2) The present invention proposes a two-stage energy storage charging and discharging strategy that takes into account the day-ahead market and the real-time market, and fully considers the economic efficiency of the energy storage project throughout its life cycle. In the day-ahead market, the energy storage kilowatt-hour cost and historical day-ahead and real-time electricity price data are used to decide whether to participate in the spot market the next day or track the wind power generation active power forecast curve, and preliminarily formulate the energy storage charging and discharging strategy in the spot market; based on the real-time electricity price prediction model of the CNN-LSTM deep neural network, and according to the real-time electricity price prediction data, by comparing the real-time electricity price and the day-ahead electricity price changes, the electricity price trend is analyzed to dynamically adjust the energy storage charging and discharging strategy set in the day-ahead market, so as to achieve the maximum energy storage project economy with the least computing power investment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of an operation method of wind power storage jointly participating in the electricity spot market based on electricity price forecasting according to the present invention;

[0044] Figure 2 This is a framework diagram of a real-time electricity price prediction model based on a CNN-LSTM deep neural network of the present invention;

[0045] Figure 3 This is a day-ahead market flow chart of the wind-storage joint operation strategy based on electricity price forecasting of the present invention;

[0046] Figure 4 This is a real-time market flow chart of the wind-storage joint operation strategy based on electricity price forecasting in the present invention. DETAILED DESCRIPTION

[0047] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0048] Example 1

[0049] The present invention proposes an operation method for wind and energy storage to jointly participate in the electricity spot market based on electricity price forecast. Consider adopting a distributed energy storage technology path in wind farms, participating in the Chinese electricity market as a "price taker" and in the form of "quoting quantity but not price". First, the energy storage system is integrated into the production management and marketing management system of the wind farm, which is suitable for the actual engineering in the initial stage of China's electricity spot market. Second, a storage charging and discharging strategy is proposed that takes into account the two stages of the day-ahead market and the real-time market, so as to achieve the maximum project economy and energy storage utilization with the least computing power investment. The specific steps are as follows: Figure 1 As shown:

[0050] Step 1: Build a real-time electricity price prediction model.

[0051] At present, the real-time electricity prices of the four moments in the previous hour are generally announced one hour later in the domestic real-time market. For example: the real-time electricity prices at 0:15, 0:30, 0:45, and 1:00 are announced during 1:00-1:15. Taking the actual situation into consideration, a deep neural network model is established with a time scale of 15 minutes, with the input being the real-time electricity prices of 96 historical moments, the output being the real-time electricity prices of the next four moments, and the sliding time window length being 4. The model uses a convolutional neural network (CNN) to extract input features, and uses the CNN extracted features as the input of a long short-term memory network (LSTM) for time series prediction, and uses the historical real-time electricity prices of the electricity spot market since its continuous operation for model training. Figure 2 As shown, the model is divided into 3 parts:

[0052] The first part consists of 5 layers of CNN, each of which is connected to a BatchNorm2d layer and a Dropout layer. The BatchNorm2d layer normalizes the data to prevent the instability of network performance due to excessive data. The Dropout layer randomly invalidates a part of the hidden layer nodes according to a certain probability in each training to prevent overfitting and improve the generalization ability of the model.

[0053] The second layer consists of two LSTM layers, which use CNN extracted features as input for time series prediction. The first layer is used as a pre-training layer. The second layer Main-LSTM is the key to improving the accuracy of time series prediction. The output result passes through a fully connected layer to adjust the vector format specified for output.

[0054] The third part is the autoregressive layer, which is parallel to the CNN-LSTM part.

[0055] Step 2: Integrate the energy storage system into the wind farm production management and marketing management system.

[0056] The automatic generation control system AGC, wind power active power forecasting system and marketing management system on the station side open data interfaces and are connected to the wind farm's energy storage management system EMS, ensuring that the energy storage system can formulate charging and discharging strategies based on wind power active power forecasting data and electricity price data.

[0057] Step 3: Develop a wind and energy storage joint operation strategy.

[0058] (3.1) Calculate the energy storage cost

[0059] Cost per kilowatt-hour The energy storage cost is calculated by leveling the cost and discharge volume over the entire life cycle of the energy storage system, as shown in formula (1).

[0060]

[0061] In the formula, the total cost of energy storage over the entire life cycle is Including the initial investment and construction costs of the project Annual operation and maintenance costs Periodic component replacement costs and cost recovery at the end of the project is the total online power consumption.

[0062]

[0063] In the formula, is the unit battery capacity of energy storage and the PCS power cost; is the energy storage capacity and the maximum charge and discharge power. is the unit battery capacity of energy storage and the maintenance cost of PCS power; r is the discount rate; Y i For the entire life cycle of energy storage. Y is the replacement cost of energy storage unit battery; i rep Replace the year collection for the energy storage battery. i is the energy storage recovery factor. is the energy storage discharge depth; N i Y The number of times the energy storage is charged annually; is the energy storage self-discharge rate; It is the rate of battery capacity decay in each energy storage cycle.

[0064] (3.2) Calculate the average ratio of the price difference between the day-ahead electricity price and the real-time electricity price

[0065] Theoretically, the day-ahead electricity price and the real-time electricity price have the same trend. However, there are situations such as fluctuations in new energy, unit tripping, and power grid failures, which lead to differences between the real-time electricity price and the day-ahead electricity price. Generally speaking, the price difference of the real-time electricity price is greater than the price difference of the day-ahead electricity price. Therefore, extract the day-ahead electricity price and real-time electricity price of historical data, calculate the difference between the average highest electricity price of h hours and the lowest electricity price of h hours, and calculate the ratio of the two.

[0066]

[0067] In the formula, It is the price difference between the day-ahead price and the average highest and lowest real-time price at h hours on a long time scale. Among them, h is the energy storage duration, usually 2 hours or 4 hours.

[0068] (3.3) Formulate a day-ahead market operation strategy

[0069] In the day-ahead market, the economic benefits of the energy storage project over its entire life cycle are fully considered and the next-day energy storage application scenario is determined based on the day-ahead electricity price. In theory, the day-ahead electricity price and the real-time electricity price have the same trend. Therefore, the specific day-ahead market operation strategy is as follows: Figure 3 As shown:

[0070] The average highest and lowest hourly and average electricity prices are calculated based on the day-ahead electricity prices at 96 times, and the day-ahead electricity price difference is further calculated.

[0071] if When the energy storage is in the spot market the next day, it will discharge and charge during the hours when the average electricity price is the highest and lowest on the previous day.

[0072] if When the day-ahead electricity price difference is small and it is not economical for energy storage to participate in the spot market, the energy storage will track the wind power generation active power forecast curve the next day to reduce the grid-connected operation assessment.

[0073] (3.4) Develop real-time market operation strategies

[0074] Due to the fluctuation of new energy, unit tripping, power grid failure and other situations, there is a difference between the real-time electricity price and the day-ahead electricity price. Therefore, by predicting the real-time electricity price and judging the trend of electricity price changes, the energy storage charging and discharging strategy is further adjusted. Specific real-time market operation strategies are as follows: Figure 4 As shown:

[0075] When the marketing platform announces the real-time electricity prices of the last four moments, it will be combined with historical data to predict the real-time electricity prices of the next four moments. Based on the predicted electricity prices, the energy storage charging and discharging strategy will be adjusted.

[0076] If the predicted real-time electricity price at time t is >1450 yuan / MWh, the energy storage is discharged for 15 minutes. If the predicted real-time electricity price at time t is <50 yuan / MWh, the energy storage is charged for 15 minutes. If the day-ahead electricity price at time t is >1450 yuan / MWh but the predicted real-time electricity price is <1000 yuan / MWh, it means that the rise of the real-time electricity price lags behind the day-ahead electricity price. If the real-time electricity price at the subsequent moment is >1450 yuan / MWh, discharge is performed. However, if there are only h hours left until 0:00 the next day, immediate discharge does not affect the energy storage operation strategy for the next day. If the day-ahead electricity price at time t is <50 yuan / MWh but the predicted real-time electricity price is >200 yuan / MWh, it means that the drop of the real-time electricity price lags behind the day-ahead electricity price. If the real-time electricity price at the subsequent moment is <50 yuan / MWh, charge is performed. However, if there are only h hours left until the discharge period, immediate charging does not affect the energy storage discharge.

[0077] Step 4: Deploy a wind and energy storage joint operation strategy.

[0078] Deploy the operation strategy to the energy storage management system EMS for operation debugging.

[0079] Example 2

[0080] As a second aspect of the present invention, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned method of wind storage jointly participating in the electricity spot market based on electricity price forecast. In addition to the above-mentioned processor, memory and interface, any device with data processing capability in which the device in the embodiment is located may also include other hardware according to the actual function of the device with data processing capability, which will not be described in detail.

[0081] Example 3

[0082] As a third aspect of the present invention, the present application also provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implements the above-mentioned method for wind and storage to jointly participate in the electricity spot market based on electricity price forecasting. The computer-readable storage medium may be an internal storage unit of any device with data processing capabilities as described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium may also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Furthermore, the computer-readable storage medium may also include both an internal storage unit and an external storage device of any device with data processing capabilities. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and may also be used to temporarily store data that has been output or is to be output.

[0083] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A method for wind power storage to jointly participate in the electricity spot market, characterized in that: The method steps include: The energy storage management system of the wind farm obtains wind power generation active power forecast data and electricity price data; In the day-ahead market, the economic feasibility of the energy storage project over its entire life cycle is considered, and the application scenarios of energy storage for the next day are determined based on the day-ahead electricity price, so as to formulate the day-ahead market operation strategy. In the real-time market, the real-time electricity price is predicted, and the trend of electricity price changes is analyzed based on the predicted real-time electricity price and the day-ahead electricity price. The energy storage charging and discharging strategy is further adjusted to formulate a real-time market operation strategy. The formulated market operation strategy is deployed to the energy storage energy management system of the wind farm for operation.

2. The method for wind power storage to jointly participate in the electricity spot market according to claim 1 is characterized in that: The day-ahead market operation strategy is formulated as follows: Calculate the average highest and lowest hourly and average electricity prices based on the day-ahead electricity prices at the set time, and further calculate the day-ahead electricity price difference If the product of the average ratio of the day-ahead electricity price to the real-time electricity price and the calculated day-ahead electricity price difference is greater than the energy storage kilowatt-hour cost, the energy storage will participate in the spot market the next day and discharge and charge during the hours when the day-ahead electricity price is the highest or lowest on average; If the product of the average ratio of the price difference between the day-ahead electricity price and the real-time electricity price and the calculated day-ahead electricity price difference is less than the energy storage cost per kilowatt-hour, the energy storage will track the wind power generation active power forecast curve the next day.

3. The method for operating a wind farm and a storage system in a joint participation in the electricity spot market according to claim 2, characterized in that: The average ratio of the price difference between the day-ahead electricity price and the real-time electricity price is calculated based on the day-ahead electricity price and the real-time electricity price of historical data, and the price difference between the average highest electricity price at h hours and the lowest electricity price at h hours is calculated, and the ratio between the two is calculated: In the formula, It is the price difference between the day-ahead electricity price and the average highest and lowest real-time electricity price at h hours on a long time scale; h is the energy storage duration.

4. The method for operating a wind farm and a storage system in a joint participation in a spot electricity market according to claim 2, characterized in that: The electricity cost The energy storage cost is calculated by leveling the cost and discharge volume over the entire life cycle of the energy storage system: In the formula, the total cost of energy storage over the entire life cycle is Including the initial investment and construction costs of the project Annual operation and maintenance costs Periodic component replacement costs and cost recovery at the end of the project is the total online power consumption; In the formula, is the unit battery capacity of energy storage and the PCS power cost; is the energy storage capacity and maximum charge and discharge power; is the unit battery capacity of energy storage and the maintenance cost of PCS power; r is the discount rate; Y i For the entire life cycle of energy storage; Y is the replacement cost of energy storage unit battery; i rep Replace the year set for the energy storage battery; i is the energy storage recovery coefficient; is the energy storage discharge depth; The number of times the energy storage is charged annually; is the energy storage self-discharge rate; It is the rate of battery capacity decay in each energy storage cycle.

5. The method for operating a wind farm and a storage system in a joint participation in the electricity spot market according to claim 1, characterized in that: The real-time market operation strategy is formulated as follows: When the marketing platform announces the real-time electricity prices of the last four moments, it combines them with historical data and predicts the real-time electricity prices of the next four moments. The energy storage charging and discharging strategy is adjusted based on the predicted electricity prices: If the predicted real-time electricity price at time t is greater than the first electricity price threshold, the energy storage will be discharged for 15 minutes; If the predicted real-time electricity price at time t is less than the second electricity price threshold, the energy storage is charged for 15 minutes.

6. The method for operating a wind farm and a storage system in a joint participation in the electricity spot market according to claim 5, characterized in that: If the day-ahead electricity price at time t is greater than the first electricity price threshold but the predicted real-time electricity price is less than the third electricity price threshold, the energy storage will not be discharged, and will be discharged when the real-time electricity price at a subsequent time is greater than the first electricity price threshold. However, if there are only h hours left until 0:00 the next day, the energy storage will be discharged immediately without affecting the energy storage operation strategy for the next day; the first electricity price threshold is greater than the third electricity price threshold; If the day-ahead electricity price at time t is less than the second electricity price threshold but the predicted real-time electricity price is greater than the fourth electricity price threshold, the energy storage will not be charged, and charging will be performed when the real-time electricity price at a subsequent moment is detected to be less than the second electricity price threshold. However, if there are only h hours left in the discharge period, immediate charging will not affect the energy storage discharge; the second electricity price threshold is less than the fourth electricity price threshold.

7. The method for wind power storage to jointly participate in the electricity spot market according to claim 1, characterized in that: The method constructs a real-time electricity price prediction model based on a CNN-LSTM deep neural network to predict the real-time electricity price; The real-time electricity price prediction model uses the historical real-time electricity prices of the electricity spot market since its continuous operation as input, uses a convolutional neural network to extract input features, and uses the CNN extracted features as the input of the long short-term memory network for time series prediction, and outputs the real-time electricity prices at the next four moments.

8. The method for operating a wind farm and a storage system in a joint participation in a spot electricity market according to claim 7, characterized in that: The real-time electricity price prediction model is as follows: The convolutional neural network part consists of 5 layers of CNN, and each CNN layer is connected to a BatchNorm2d layer and a Dropout layer; the BatchNorm2d layer normalizes the data; the Dropout layer randomly invalidates a part of the hidden layer nodes according to a certain probability in each training; The long short-term memory network consists of two layers of LSTM. The features extracted by the convolutional neural network are used as input for time series prediction. The first layer is used as a pre-training layer, and the second layer is the Main-LSTM. The output results pass through a fully connected layer to adjust the vector format specified for the output. Autoregressive layer, in parallel with the CNN-LSTM part.

9. An operating device for wind power storage to jointly participate in the electricity spot market, comprising a memory, a processor, and a program stored in the memory, characterized in that: When the processor executes the program, it implements the operating method of wind power storage jointly participating in the electricity spot market as described in any one of claims 1-8.

10. A storage medium having a program stored thereon, characterized in that: When the program is executed, the operating method of wind and energy storage jointly participating in the electricity spot market as described in any one of claims 1-8 is implemented.

Images