Method for enabling energy storage system to cooperate with wind power station to participate in electric quantity bidding in electric power market

Abstract

A method for an energy storage system to cooperate with a wind power station to participate in electricity market electricity bidding relates to the field of electricity market electricity management. The invention aims to solve the problem that the economic benefit of the wind storage power station is influenced because the existing bidding strategy of the wind storage power station cannot comprehensively and deeply consider the important regulation and control influence of the electricity price fluctuation on the declaration electric quantity. The method comprises the following steps: calculating a historical electricity price standard deviation and an average value by using historical electricity prices in a historical time period; predicting a predicted electricity price in a future time period by using the electricity price prediction model; when the predicted electricity price is larger than the sum of the historical electricity price standard deviation and the average value, an upper limit bidding strategy is adopted, and when the predicted electricity price is larger than the sum of the historical electricity price standard deviation and the average value and smaller than the difference of the historical electricity price standard deviation and the average value, the upper limit bidding strategy is adopted; and adopting a full-electric-quantity bidding strategy, and adopting a lower-limit bidding strategy when the predicted electricity price is smaller than the difference between the historical electricity price standard deviation and the average value.

Description

Technical Field

[0001] The invention belongs to the field of power market electricity quantity management. Background Art

[0002] As a pollution-free and renewable new energy, wind energy has attracted much attention among many clean energy sources. However, its inherent randomness and volatility mean that wind power generation is restricted by multiple natural factors such as wind speed, temperature, and rainfall, which has a significant impact on power generation and further affects power quality and grid stability.

[0003] In order to meet this challenge, the introduction of energy storage devices has become a recognized solution in the industry. When wind power generation exceeds the immediate demand of the power grid, the energy storage system can absorb and store excess electricity; on the contrary, when wind power generation is insufficient to meet the needs of the power grid, the energy storage system releases the stored electricity to make up for the gap, thereby greatly improving the utilization efficiency and economic benefits of new energy.

[0004] With the acceleration of the process of electricity marketization, the deep integration of energy storage technology and wind power generation has become an irreversible trend. This integration strategy not only helps to increase the proportion of renewable energy in the overall energy structure, but also effectively overcomes the problem of large fluctuations and instability of wind energy. In the power market environment, formulating and implementing a robust bidding strategy is crucial to enhancing the market competitiveness of wind power. By actively participating in the power market bidding, wind storage power stations can use energy storage facilities more efficiently, flexibly adjust wind power generation, and achieve agile response to power market demand and optimal allocation of resources. This can not only promote the sustainable development of the wind power industry, but also effectively reduce the backup burden of the power system and improve the reliability and economy of power supply. Therefore, in-depth exploration of the regulation strategy of wind storage power stations in the power spot market has far-reaching significance for guiding the operation of the future power market and the integration of new energy.

[0005] The existing bidding strategy for wind-storage power stations still has defects and fails to fully and deeply consider the important regulatory impact of electricity price volatility on the declared electricity volume. This makes wind-storage power stations face greater market risks and uncertainties in the bidding process, which in turn has a significant negative impact on their economic benefits. Summary of the invention

[0006] The present invention aims to solve the problem that the existing bidding strategy for wind power stations fails to fully and deeply consider the important regulatory impact of electricity price volatility on the declared electricity volume, thereby affecting the economic benefits of wind power stations. A method for an energy storage system to cooperate with a wind power station to participate in the bidding for electricity volume in the electricity market is now provided.

[0007] The method of energy storage system cooperating with wind power station to participate in electricity market electricity bidding includes:

[0008] Calculate the standard deviation of historical electricity prices using historical electricity prices within the historical period history and the average value μ history ;

[0009] Use the electricity price forecasting model to predict the predicted electricity price λ for the future period forecast ;

[0010] According to the predicted electricity price λ forecast The standard deviation of historical electricity prices history and the average value μ history The relationship determines the bidding strategy:

[0011] When forecast >μ history +σ history When the upper limit bidding strategy is adopted,

[0012] When μ history -σ history ≤λ forecast ≤μ history +σ history When the full electricity bidding strategy is adopted,

[0013] When forecast <μ history -σ history When the bid is less than 1%, the lower limit bidding strategy is adopted.

[0014] Furthermore, under the above upper limit bidding strategy, the expression of the current bidding quantity Q is:

[0015] Q=(1+x%)Q0,

[0016] Under the full electricity bidding strategy, the expression of the current bidding electricity Q is:

[0017] Q=Q0,

[0018] Under the lower limit bidding strategy, the expression of the current bidding quantity Q is:

[0019] Q = (1-x%) Q0,

[0020] In the above formula, Q0 is the basic bid electricity, and x% is the preset electricity adjustment ratio.

[0021] Furthermore, the expression of the above basic bidding quantity Q0 is:

[0022] Q0=P t bid Δt=(P t w +P t dch -P t ch )Δt,

[0023] Among them, Δt is the sampling step at the time point, P t bid is the control power at the current time t, P t w is the wind power station power output decision at the current time t, P t dch is the energy storage discharge power at the current time t, P t ch is the energy storage charging power at the current time t.

[0024] Furthermore, the method for selecting the above preset power adjustment ratio x% is:

[0025] According to the predicted electricity price λ in the future period forecast Calculate its standard deviation σ forecast ,

[0026] If σ forecast >σ history , then the preset power adjustment ratio x% is the conservative power adjustment ratio.

[0027] If σ forecast ≤σ history , then the preset power adjustment ratio x% is the aggressive power adjustment ratio.

[0028] Furthermore, the training method of the above electricity price prediction model includes:

[0029] Establish a wind power data feature set and a wind power data output set, wherein the samples in the wind power data feature set include: historical electricity price data of the power market for N consecutive hours, historical power load samples of the wind storage power station for N consecutive hours, historical wind speed samples of the wind storage power station for N consecutive hours, and historical power generation samples of the wind storage power station for N consecutive hours, and the samples in the wind power data output set include: electricity price data for M consecutive hours;

[0030] The samples in the wind power data feature set are used as input, and the samples in the wind power data output set are used as output, a model capable of converting multiple feature quantities into a single feature quantity is constructed, and the model parameters are trained to obtain an electricity price prediction model.

[0031] Furthermore, the above electricity price prediction model is an ICA-Timesnet prediction model.

[0032] Furthermore, the profit objective function of the wind power station and the spot market is:

[0033]

[0034] Where T is the total number of sampling time points, For the imbalance penalty fee, Battery loss fee.

[0035] Furthermore, the above imbalance penalty fee The expression is:

[0036]

[0037] Among them, m pun is the penalty coefficient for unbalanced electricity, λ is the regulated electricity price, P t da To regulate the power consumption on the day before, P t rt To control power in real time, P t w is the wind power station power output decision at the current time t, P t dch is the energy storage discharge power at the current time t, P t ch is the energy storage charging power at the current time t;

[0038] The battery loss cost The expression is:

[0039]

[0040] Among them, c0 is the unit cost of charging and discharging, and Δt is the sampling step at the time point.

[0041] Furthermore, the constraints for the energy storage system to cooperate with the wind power station to participate in the electricity market electricity bidding include energy storage constraints and charge state constraints.

[0042] Furthermore, the above energy storage constraint is:

[0043] When the energy storage system is discharged:

[0044] When the energy storage system is charging:

[0045] Among them, P t dch is the energy storage discharge power at the current time t, P t ch is the energy storage charging power at the current time t, is the maximum discharge power of energy storage, is the maximum charging power of energy storage, is the energy storage discharge state variable, is the energy storage charging state variable, and:

[0046] The state of charge constraint is:

[0047] SOC t =W t / W max ,

[0048] SOC min ≤SOC t ≤SOC max ,

[0049] In the formula, SOC t is the state of charge of the energy storage system at time t, W t is the power of the energy storage system at the end of time t, W max is the maximum capacity of the energy storage system, SOC min is the lower limit of the energy storage system, SOC max is the upper limit of the energy storage system.

[0050] The beneficial effects of the present invention are as follows:

[0051] By comparing the standard deviation of historical electricity prices with predicted electricity prices, the range of bid electricity can be dynamically adjusted to maximize the profitability of new energy power stations while controlling risks.

[0052] The method of the present invention uses historical electricity price data and adopts the network time series model ICA-Timesne based on independent component analysis to predict electricity prices, effectively capturing the pattern and trend of electricity price fluctuations, thereby providing more accurate prediction results. By comparing the standard deviation of historical electricity prices and predicted electricity prices, the size of electricity price fluctuations is judged, and appropriate bidding strategies are selected to help wind power stations adopt the most appropriate risk management methods in different market environments to ensure economic maximization.

[0053] By combining electricity price forecasting, bidding strategies and energy storage systems, the present invention enables wind farms to more accurately grasp electricity price trends in the day-ahead market, reduce the impact of uncertainty, and thus improve market competitiveness. By adopting appropriate upper and lower limits for bidding, wind farms can not only effectively avoid market risks, but also maximize their profits from wind power generation. In addition, the reasonable scheduling of the energy storage system enables wind farms to maintain a relatively stable income stream even when electricity prices fluctuate greatly, ensuring the maximization of the economic benefits of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flow chart of the method for energy storage system to cooperate with wind power station to participate in electricity market electricity bidding;

[0055] Figure 2 It is the flow chart of ICA-Timesnet electricity price forecasting model;

[0056] Figure 3It is the electricity price forecasting results and evaluation indicators using the ICA-Timesnet electricity price forecasting model;

[0057] Figure 4 It is the 96-point bidding power ratio and profit curve of the conservative strategy;

[0058] Figure 5 It is the 96-point bidding power ratio and profit curve of the aggressive strategy;

[0059] Figure 6 It is the profit comparison result of full electricity bidding, conservative bidding, aggressive bidding and the bidding strategy of this paper. DETAILED DESCRIPTION

[0060] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0061] This implementation method constructs a detailed bidding strategy by taking into account the uncertainty of output and market price faced by wind farms equipped with energy storage systems when participating in the day-ahead market bidding. Figure 1 and Figure 2 Specifically describing this implementation mode, the method described in this implementation mode for the energy storage system to cooperate with the wind power station to participate in the bidding of electricity quantity in the electricity market includes:

[0062] In this implementation, the electricity price standard deviation is introduced to determine the magnitude of electricity price fluctuations and select an appropriate bidding strategy.

[0063] 1. Calculate the standard deviation of historical electricity prices based on the electricity price data of 96 time points one day before the current moment history With the average value μ history , as a benchmark for subsequent analysis.

[0064] The ICA-Timesnet forecasting model is used as the electricity price forecasting model considering the characteristics of wind power data. The electricity price data within one week before the current moment is used for modeling. The electricity prices at 96 time points in the next day after the current moment are predicted, and the standard deviation σ of these predicted electricity prices is calculated. forecast .

[0065] By comparing the standard deviation σ of historical electricity prices history The standard deviation of the predicted electricity price σ forecast The size of is used to determine the upper and lower limits of the bidding strategy. Specifically:

[0066] If σ forecast >σ history , then the predicted electricity price volatility is large and the bidding risk is high. At this time, a conservative bidding strategy should be adopted. The bidding range of electricity quantity should be artificially narrowed according to actual demand to avoid bidding risk losses and choose a conservative electricity quantity adjustment ratio a.

[0067] If σ forecast ≤σ history , then the predicted electricity price volatility is small and the bidding risk is low. At this time, an aggressive bidding strategy is adopted. The electricity bidding range should be artificially expanded according to actual demand to improve the economic benefits of the wind storage power station and select the aggressive electricity adjustment ratio b.

[0068] 2. Based on the predicted electricity price λ forecast Compared with the historical average electricity price μ history and standard deviation σ history The relationship between electricity prices and electricity prices can be further determined, and different bidding strategies can be adopted accordingly.

[0069] When forecast >μ history +σ history When the electricity price is considered high, the upper limit bidding strategy is adopted to ensure that the energy storage system can supplement the insufficient electricity and sell the energy storage system's energy storage as much as possible. The current bidding power Q = (1 + x%) Q0.

[0070] When μ history -σ history ≤λ forecast ≤μ history +σ history When the electricity price is considered normal, the full electricity bidding strategy is adopted, and the current bidding electricity Q = Q0;

[0071] When forecast <μ history -σ history When , the electricity price is considered low, and the lower limit bidding strategy is adopted. The current bid power Q = (1-x%) Q0, and the excess power is stored in the energy storage system.

[0072] The above x% is the preset electricity adjustment ratio. When a conservative bidding strategy is adopted, x%=a, and when an aggressive bidding strategy is adopted, x%=b, and b>a; Q0 is the basic bid electricity.

[0073] Through this series of strategy formulation and execution, the flexibility and profitability of wind farms in participating in day-ahead market bidding can be effectively improved. The effectiveness and feasibility of this method can be verified through actual case analysis. The implementation of this implementation will effectively improve the effectiveness and feasibility of wind farms equipped with energy storage systems participating in power market regulation.

[0074] In this implementation, the profit objective function of the wind power station and the spot market is:

[0075]

[0076] Where T represents the total number of sampling time points, t = 1, 2, ..., T;

[0077] represents the imbalance penalty fee, which is expressed as:

[0078]

[0079] Represents the battery loss cost, expressed as:

[0080]

[0081] In the above formula, m pun represents the penalty coefficient of unbalanced electricity, λ represents the regulated electricity price, P t da Indicates the amount of electricity regulated the day before, P t rt represents real-time control of power, c0 represents the unit cost of charging and discharging, Δt is the sampling step at the time point, P t w is the wind power station power output decision at the current time t, P t dch is the energy storage discharge power at the current time t, P t ch is the energy storage charging power at the current time t.

[0082] In this implementation, in order to ensure that the power control can accurately reflect the actual situation, the basic bidding power Q0 is set to the sum of the wind power plant power generation and the energy storage system charge and discharge capacity:

[0083] Q0=P t bid Δt=(P t w +P t dch -P t ch )Δt,

[0084] Where P t bid is the control power at the current time t.

[0085] In order to ensure the safe and stable operation of renewable energy power generation, the decision output power P of renewable energy power generation t w is strictly limited and must maintain the predicted power P of the wind power station at the current time tt wf Less than or equal to the predicted power, that is, 0≤P t w ≤P t wf To this end, the following specific constraints are set:

[0086] 1) Energy storage constraints

[0087] In order to more efficiently integrate and balance the fluctuating output characteristics of wind power generation and to extend the service life of the energy storage system, this implementation method comprehensively considers various characteristics of the energy storage system itself, including its current charging state, the maximum allowable charging and discharging amount, and the charging and discharging power limit.

[0088] The maximum charge and discharge power of the energy storage system at time t is constrained by its remaining power and its own characteristics. These characteristics jointly determine the maximum power value that the energy storage system can charge and discharge at a given time point. Therefore, the energy storage discharge power P at the current time t is t dch and the energy storage charging power P at the current time t t ch The constraints are:

[0089] When the energy storage system is discharged:

[0090] When the energy storage system is charging:

[0091] in, is the maximum discharge power of energy storage, is the maximum charging power of energy storage, is the energy storage discharge state variable, is the energy storage charging state variable, and:

[0092]

[0093] 2) State of charge constraints

[0094] In order to optimize the performance of the energy storage system during charging and discharging, and at the same time reduce the adverse effects of the state of charge on the system, corresponding constraints must be imposed on it. Therefore, the state of charge constraints of the energy storage system are:

[0095] SOC t =W t / W max ,

[0096] SOC min ≤SOC t ≤SOC max ,

[0097] In the formula, SOC t is the state of charge of the energy storage system at time t, W t Represents the amount of electricity in the energy storage system at the end of time t, W max Indicates the maximum capacity of the energy storage system, SOC min Indicates the lower limit of the energy storage system, SOC max Indicates the upper limit of the energy storage system.

[0098] The amount of energy storage system at the end of time t+1 is W t+1 for:

[0099] W t+1 =W t +P t ch η ch Δt-P t dch η dch Δt;

[0100] Where η ch is the charging efficiency, η dch is the discharge efficiency.

[0101] In this implementation, the training method of the electricity price prediction model is:

[0102] Establish a wind power data feature set and a wind power data output set; the training samples in the wind power data feature set include: historical electricity price data of the power market for N consecutive hours, historical power load samples of the wind storage power station for N consecutive hours, historical wind speed samples of the wind storage power station for N consecutive hours, and historical power generation samples of the wind storage power station for N consecutive hours; the wind power data output set includes electricity price data for M consecutive hours.

[0103] The electricity price prediction model performs independent component analysis on the wind power data feature set and converts multiple feature quantities into a single feature quantity. Based on the single feature quantity, the Timesnet prediction model is used to predict the historical electricity price data. By adjusting the network parameters of the electricity price prediction model, the predicted electricity price for the electricity bidding strategy is finally obtained.

[0104] The output result of the electricity price prediction model is the day-ahead electricity price for the next M hours. If a sampling point is taken at an interval of 15 minutes, a total of 4M sampling points will be formed for M consecutive hours.

[0105] Common objective evaluation indicators of electricity price forecasting models include root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and coefficient of determination (R 2 ) can be used to fully evaluate the performance of the algorithm.

[0106]

[0107] In the formula, y i represents the actual value of the corresponding electricity price of the i-th sample, represents the corresponding electricity price forecast value of the i-th sample, Represents the average value of the actual value of the corresponding electricity price of the i-th sample. The smaller the MAE, RMSE and MAPE indicators are, the better the prediction effect of the model is. 2 It measures how well the model explains the variability in the data, and the closer it is to 1, the better the model.

[0108] Verification test:

[0109] In order to verify the effect of this implementation method, a wind power station with energy storage equipment was selected and a simulation analysis was performed. The specific data are as follows: the total installed capacity of the wind power station is 220MW; the deviation exemption margin is 5%; the deviation penalty coefficient is set to 0.5; the capacity of the energy storage equipment is 40MW / 80MWh; the charging and discharging efficiency is 0.8; the unit operating cost is 24 yuan / (MWh); the initial state of charge is 0.5. The wind power output data comes from the data set of the actual wind power base, and the spot electricity price data is taken from the day-ahead electricity price data set of the power market.

[0110] Regarding the bidding strategy, three strategies are used for comparison, namely:

[0111] 1) The bidding strategy proposed in this implementation method is: based on the comparison of standard deviations, different bidding upper and lower limits are selected. If the historical standard deviation is smaller than the predicted standard deviation, a conservative bidding strategy is adopted, and the bidding upper and lower limits are 90% to 110%; if the historical standard deviation is larger than the predicted standard deviation, an aggressive bidding strategy is adopted, and the bidding upper and lower limits are 60% to 140%.

[0112] 2) Conventional bidding ratio strategy: adopt the electricity adjustment ratio that complies with the regulations of the electricity market, with the upper and lower limits of the bidding being 120% and 80% respectively.

[0113] 3) Full electricity declaration strategy: 100% bidding is conducted based on the power generation of the wind power station.

[0114] Figure 3The electricity price forecast results of the ICA-Timesnet forecasting model are shown. The predicted values ​​of the forecasting model are basically consistent with the actual values. Among them, RMSE measures the square root of the average deviation between the predicted value and the actual value. A smaller RMSE value indicates that the model prediction error is smaller. The RMSE value of the forecast result is 2.744, indicating that the prediction error of the model is in a reasonable range and the error level is low. MAE measures the average absolute error between the predicted value and the actual value. The smaller the value, the more accurate the model prediction. The MAE of the forecast result is 2.149, which means that the average absolute error between each predicted value and the true value is 2.149. The error is not large and can be considered a relatively small error. MAPE reflects the percentage of model prediction error to the actual value. The lower the better. A MAPE value of 3.973% means that the model's prediction error is only 3.973% of the actual value. In practical applications, this error is very small, indicating that the model's prediction accuracy is high. R 2 It measures the proportion of the true value variation that the model can explain. The closer the value is to 1, the better the model fits the data, and the R 2 It is 0.969, which is close to 1, indicating that the model can explain 96.9% of the data variation and the fitting effect is very good. Overall, the ICA-Timesnet prediction model used in this experiment not only has a small error, but also performs well in accuracy and fit.

[0115] Figure 4 The bid volume curve and profit curve when the conservative bidding strategy is adopted are shown. When the historical standard deviation is less than the predicted standard deviation, it indicates that the electricity price fluctuates greatly, and we choose the conservative bidding strategy. Under this strategy, the upper and lower limits of the bid are set at 90% to 110%. Figure 4 It can be seen that in the 0-22 period, the lower limit bidding was adopted, the profit was low, and the profit curve showed a downward trend. In the 37-51 and 62-84 period, the upper limit bidding was adopted, the profit increased, and the profit curve rose.

[0116] Figure 5 The bid volume curve and profit curve when the aggressive bidding strategy is adopted are shown. When the historical standard deviation is greater than the predicted standard deviation, it means that the electricity price fluctuation is small and an aggressive bidding strategy is adopted. Under this strategy, the upper and lower limits of the bid are set at 75% to 125%. Figure 5 It shows that in the time period of 0-22, the lower limit bidding was also adopted, the profit was low, and the profit curve fell slightly. In the time period of 64-82, the upper limit bidding was adopted, the profit increased, and the profit curve rose accordingly.

[0117] By comparison Figure 4 (b) with Figure 5(b), it can be observed that the profit upper bound of the aggressive bidding strategy is higher, while the lower bound is lower, which is consistent with the characteristics of the aggressive bidding strategy.

[0118] Figure 6 The profit comparison result diagram of the method of this embodiment and the other two strategies for regulation is shown. Deviation penalty reflects the risk brought by electricity price fluctuations. Lower deviation penalty means that the bidding strategy can better avoid the risk of electricity price fluctuations, thereby reducing the economic losses caused by electricity price fluctuations. In this regard, the 4473.344 yuan of the bidding strategy of this embodiment is significantly lower than the 5638.249 yuan of the conventional multiple bidding strategy, and is close to the 3259.679 yuan of the full power bidding strategy, which shows that the bidding strategy of the present invention performs better in avoiding excessive deviations. Battery loss is related to the efficiency of the battery energy storage system and the rationality of the bidding strategy. The smaller the loss, the better. From the data in the table, it can be seen that although the battery loss of the bidding strategy of this embodiment is slightly higher than that of the full power bidding strategy, it is significantly lower than the conventional multiple bidding strategy. This shows that the bidding strategy of this embodiment is more effective than the conventional multiple bidding strategy in reducing battery loss. Although the bidding profit of the conventional multiple bidding strategy is slightly higher, the profit of the bidding strategy of this embodiment is close to the conventional strategy and is higher than the full power bidding strategy. Taking into account other indicators (such as deviation penalty and battery loss), this shows that the bidding strategy of this implementation method is stable in terms of bidding profit and can provide better overall benefits. In terms of the total profit indicator, the bidding strategy of this implementation method performs better than the other two strategies, especially compared with the conventional multiple bidding strategy, the total profit of this implementation method strategy is higher, reaching 115,382.645 yuan. This shows that although the strategy of this implementation method is similar to or slightly lower than the conventional strategy in some individual indicators, it can bring higher overall economic benefits after comprehensive consideration. Therefore, the bidding strategy of this implementation method is undoubtedly a better bidding strategy. It not only performs well in economic benefits, but also has obvious advantages in reducing deviations and battery losses, which meets the requirements for efficiency and sustainability of bidding strategies in the power market.

[0119] In summary, the existing bidding strategy for wind-storage power stations fails to fully consider the regulatory impact of electricity price volatility on the bid electricity, resulting in the wind-storage power stations facing greater risks in the bidding process, which in turn affects their economic benefits. The present invention provides an improved electricity bidding strategy that effectively avoids the risk of electricity price fluctuations, maximizes the economic benefits of wind-storage power stations, and improves the economy of the bidding strategy. The present invention is applicable to the electricity bidding of wind-storage power stations in the electricity market to optimize the economic benefits of power stations.

[0120] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in a manner different from that described in the original claims. It should also be understood that the features described in conjunction with a single embodiment may be used in other described embodiments.

Claims

1. A method for an energy storage system to cooperate with a wind power station to participate in bidding for electricity in the electricity market, characterized in that: include: Calculate the standard deviation of historical electricity prices using historical electricity prices within the historical period history and the average value μ history ; Use the electricity price forecasting model to predict the predicted electricity price λ for the future period forecast ; According to the predicted electricity price λ forecast The standard deviation of historical electricity prices history and the average value μ history The relationship determines the bidding strategy: When forecast >μ history +σ history When the upper limit bidding strategy is adopted, When μ history -σ history ≤λ forecast ≤μ history +σ history When the full electricity bidding strategy is adopted, When forecast <μ history -σ history When , a lower limit bidding strategy is adopted.

2. The method for using an energy storage system in conjunction with a wind power station to participate in bidding for electricity in the electricity market according to claim 1, characterized in that: Under the upper limit bidding strategy, the expression of the current bidding quantity Q is: Q=(1+x%)Q0, Under the full electricity bidding strategy, the expression of the current bidding electricity Q is: Q=Q0, Under the lower limit bidding strategy, the expression of the current bidding quantity Q is: Q = (1-x%) Q0, In the above formula, Q0 is the basic bid electricity, and x% is the preset electricity adjustment ratio.

3. The method for using an energy storage system in conjunction with a wind power station to participate in bidding for electricity in the electricity market according to claim 2, characterized in that: The expression of the basic bidding quantity Q0 is: Q0=P t bid Δt=(P t w +P t dch -P t ch )Δt, Among them, Δt is the sampling step at the time point, P t bid is the control power at the current time t, P t w is the wind power station power output decision at the current time t, P t dch is the energy storage discharge power at the current time t, P t ch is the energy storage charging power at the current time t.

4. The method for participating in the bidding of electricity quantity in the electricity market by using an energy storage system in cooperation with a wind power station according to claim 2 or 3, characterized in that: The method for selecting the preset power adjustment ratio x% is: According to the predicted electricity price λ in the future period forecast Calculate its standard deviation σ forecast , If σ forecast >σ history , then the preset power adjustment ratio x% is the conservative power adjustment ratio. If σ forecast ≤σ history , then the preset power adjustment ratio x% is the radical power adjustment ratio.

5. The method for using an energy storage system in conjunction with a wind power station to participate in bidding for electricity in the electricity market according to claim 4, characterized in that: The training method of the electricity price prediction model comprises: Establish a wind power data feature set and a wind power data output set, wherein the samples in the wind power data feature set include: historical electricity price data of the power market for N consecutive hours, historical power load samples of the wind storage power station for N consecutive hours, historical wind speed samples of the wind storage power station for N consecutive hours, and historical power generation samples of the wind storage power station for N consecutive hours, and the samples in the wind power data output set include: electricity price data for M consecutive hours; The samples in the wind power data feature set are used as input, and the samples in the wind power data output set are used as output, a model capable of converting multiple feature quantities into a single feature quantity is constructed, and the model parameters are trained to obtain an electricity price prediction model.

6. The method for participating in the bidding of electricity quantity in the electricity market by using an energy storage system in cooperation with a wind power station according to claim 1 or 5, characterized in that: The electricity price prediction model is the ICA-Timesnet prediction model.

7. The method for using an energy storage system in conjunction with a wind power station to participate in bidding for electricity in the electricity market according to claim 2, characterized in that: The profit objective function of wind power station and spot market is: Where T is the total number of sampling time points, For the imbalance penalty fee, Battery loss fee.

8. The method for using an energy storage system in conjunction with a wind power station to participate in bidding for electricity in the electricity market according to claim 7, characterized in that: The imbalance penalty fee The expression is: Among them, m pun is the penalty coefficient for unbalanced electricity, λ is the regulated electricity price, P t da To regulate the power consumption on the day before, P t rt To control power in real time, P t w is the wind power station power output decision at the current time t, P t dch is the energy storage discharge power at the current time t, P t ch is the energy storage charging power at the current time t; The battery loss cost The expression is: Among them, c0 is the unit cost of charging and discharging, and Δt is the sampling step at the time point.

9. The method for using an energy storage system in conjunction with a wind power station to participate in bidding for electricity in the electricity market according to claim 1, characterized in that: The constraints for energy storage systems to cooperate with wind power stations in participating in electricity market electricity bidding include energy storage constraints and charge state constraints.

10. The method for participating in the bidding of electricity quantity in the electricity market by using an energy storage system in cooperation with a wind power station according to claim 9, characterized in that: The energy storage constraint is: When the energy storage system is discharged: When the energy storage system is charging: Among them, P t dch is the energy storage discharge power at the current time t, P t ch is the energy storage charging power at the current time t, is the maximum discharge power of energy storage, is the maximum charging power of energy storage, is the energy storage discharge state variable, is the energy storage charging state variable, and: The state of charge constraint is: SOC t =W t / W max , SOC min ≤SOC t ≤SOC max , In the formula, SOC t is the state of charge of the energy storage system at time t, W t is the power of the energy storage system at the end of time t, W max is the maximum capacity of the energy storage system, SOC min is the lower limit of the energy storage system, SOC max is the upper limit of the energy storage system.

Images