Wind power plant price random optimization method and system based on wind speed uncertainty

By generating different wind speed scenarios and establishing a frequency response scheduling cost model, the problem of evaluating the operating costs of the power system by wind power uncertainty is solved, and more accurate wind farm price evaluation and power scheduling are achieved, reducing calculation errors and time.

CN119090536APending Publication Date: 2024-12-06SHENYANG UNIVERSITY OF TECHNOLOGY +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411271918.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-11
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The prior art is difficult to effectively evaluate and alleviate the impact of wind power uncertainty on the operating costs of power systems, especially in high time resolution and a large number of representative scenarios, resulting in insufficient price assessment of the power market.

Method used

The wind farm price stochastic optimization method based on wind speed uncertainty is adopted to generate wind power generation scenarios without wind speed and with wind speed uncertainty by collecting historical wind speed data, establish a frequency response scheduling cost model, and calculate the power generation cost difference in each time segment. Frequency response deployment cost differences and wind energy cut penalty cost differences are used to determine the time-sharing price of wind farms.

Benefits of technology

It reduces prediction errors, reduces power scheduling costs, and improves the real-time and accuracy of calculations, which can more accurately characterize the impact of wind speed uncertainty on wind farm prices.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119090536A_ABST
    Figure CN119090536A_ABST
Patent Text Reader

Abstract

The invention discloses a wind power plant price random optimization method and system based on wind speed uncertainty, and the method comprises the steps: collecting the historical wind speed data of a wind power plant, and generating a wind power generation scene without wind speed uncertainty and a wind power generation scene with wind speed uncertainty; establishing a frequency response scheduling cost model; based on a frequency response scheduling cost model, according to the operation data in the wind power generation scene without wind speed uncertainty and the operation data in the wind power generation scene with wind speed uncertainty, the frequency response scheduling cost model is scheduled according to the operation data in the wind power generation scene without wind speed uncertainty. Respectively determining a power generation cost difference, a frequency response deployment cost difference and a wind energy reduction punishment cost difference corresponding to each time segment in the two wind power generation scenes; the sum of the power generation cost difference, the frequency response deployment cost difference and the wind energy reduction punishment cost difference corresponding to the time segments in the two wind power generation scenes is used as the time-sharing price of the wind power plant, and the wind power uncertainty price is evaluated through high time resolution and a large number of representative scenes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of new energy technology, and specifically relates to a method for uncertainty pricing based on stochastic optimization, especially to a method for uncertainty pricing based on stochastic optimization for cost calculation of frequency response (FR) scheduling of wind power. Background Art

[0002] Due to the widespread use of renewable energy such as wind power, its risks have also brought challenges to the operation of the power system. If not handled properly, it may damage economic benefits and have a negative impact on the reliability of the system. The risks of renewable energy mainly include two aspects: variability and uncertainty.

[0003] To eliminate the impact of uncertainty, it is necessary to first forecast renewable energy output. In order to reduce errors and improve accuracy, existing forecasting methods include: physical methods, which mainly use physical data to generate weather and wind forecasts over a period of time. Physical methods include using spatial correlation models to exploit the spatial relationship between different wind speed data and physical properties, and predicting the wind speed at these locations by analyzing the patterns and important parameters of the data obtained at specific locations. Statistical methods are used to analyze data, and the patterns of future wind speed and wind energy output are depicted based on probability density functions (PDFs); Monte Carlo simulations based on probability density functions create wind energy output scenarios, allowing power system operators to take uncertainty into account. Statistical methods are easy to implement, but their prediction errors are relatively large as the forecast time increases. Artificial intelligence algorithms use historical data to detect complex nonlinear relationships and determine the dependencies between different variables that affect wind speed forecasts with high accuracy. In order to eliminate the impact of wind power uncertainty on wind power pricing, a flexible wind power security constraint framework coordinates supply and demand in an appropriate manner to meet the demand for greater flexibility in a safe, economical and green power grid, and on this basis determines the optimal time-of-use electricity price. The uncertainty of wind power generation will have a significant impact on the price of wind power transactions, and the time-of-use electricity price mechanism is more in line with the characteristics of wind power transactions. In order to reduce the impact of wind power uncertainty on wind power pricing, time-of-use electricity prices or fixed-area pricing methods predict wind power uncertainty from a large-scale time or space, and cannot accurately price the wind power of each wind farm, and there are still large errors and other problems.

[0004] Since renewable energy has almost no frequency regulation capability, its weak regulation characteristics require high dispatching capability of the power system. Therefore, when a large amount of renewable energy is incorporated into power systems such as power grids, the operation and dispatching costs of the power system will increase. Existing technologies use FR to improve the regulation capability of renewable energy. Taking wind power as an example, FR means allowing generators to quickly reduce power generation when wind power supply increases and increase power generation when wind power supply decreases. It provides the flexibility required by the power system to adapt to the uncertainty of renewable energy resources. The frequency response dispatch method includes a piecewise linear simplified model of the frequency response process of conventional units, and a corresponding optimal power flow model is constructed to improve the frequency regulation capability of energy; the inertial response of wind power (i.e., the weak regulation characteristic of wind power) is equivalent to a fixed inertial time constant, and the benefits of wind power frequency support are studied in the dispatching framework. Due to the integration of renewable energy, as the purchase of FR increases, the operating cost of the power system will inevitably increase.

[0005] Existing technologies focus on the impact of wind power uncertainty on the power market, analyze the impact of wind power on operating reserves and the impact of wind power uncertainty on electricity prices from the perspective of unit commitment and market, and use robust optimization or stochastic optimization to mitigate the impact of wind power uncertainty. Although there have been studies on ancillary service scheduling that consider wind power uncertainty, there is still a gap in the impact of wind power uncertainty on the increase in power system operating costs, and there is a lack of a comprehensive method to evaluate the price of wind power uncertainty with high temporal resolution and a large number of representative scenarios. Summary of the invention

[0006] In order to solve the deficiencies in the prior art, the present invention provides a random optimization method and system for wind farm prices based on wind speed uncertainty. With the continuous improvement of wind energy penetration, the demand for quantifying the impact of wind energy uncertainty based on the power system optimization model of FR in the random optimization framework is growing, so as to evaluate the impact of wind power uncertainty on the power market. The present invention divides the wind power uncertainty price into two parts. One part is the wind power uncertainty price caused by the change in power generation due to wind uncertainty. The frequency response scheduling cost introduced in the power dispatching process to ensure the regulation characteristics is also a part of the wind power uncertainty price. The wind power uncertainty price is evaluated with high time resolution and a large number of representative scenarios.

[0007] The present invention adopts the following technical solution.

[0008] The present invention proposes a stochastic optimization method for wind farm prices based on wind speed uncertainty, comprising:

[0009] Collect historical wind speed data of wind farms to generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty;

[0010] Establish a frequency response dispatch cost model;

[0011] Based on the frequency response dispatch cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, the power generation cost difference, frequency response deployment cost difference and wind energy curtailment penalty cost difference corresponding to each time segment under the two wind power generation scenarios are determined respectively;

[0012] The sum of the power generation cost difference, frequency response deployment cost difference and wind energy reduction penalty cost difference corresponding to each time segment under the two wind power generation scenarios is taken as the time-of-use price of the wind farm.

[0013] Preferably, a stochastic optimization model is established, wherein the stochastic optimization model takes historical wind speed data of the wind farm as input and takes wind power generation scenarios as output;

[0014] The stochastic optimization model has multiple parallel scenario generation layers built in, and each scenario generation layer outputs the corresponding wind power generation scenario based on different probability distribution models;

[0015] The stochastic optimization model is used to generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty.

[0016] Preferably, the stochastic optimization model is used to generate a wind power generation scenario without wind speed uncertainty and a wind power generation scenario with wind speed uncertainty, including:

[0017] The collected historical wind speed data are sampled and the sampling values ​​are verified using the time series sampling method;

[0018] The wind speed mean and standard deviation of each time interval are calculated based on the sampled and verified historical wind speed data; the corresponding wind power generation scenarios are generated using multiple scenario generation layers based on the wind speed mean and standard deviation of each time interval;

[0019] A single wind speed scenario is used as a wind power generation scenario without wind speed uncertainty, and multiple wind speed scenarios constitute a wind power generation scenario with wind speed uncertainty.

[0020] Preferably, establishing a frequency response dispatch cost model includes:

[0021] Divide a day into multiple time segments, and establish a power system cost model for each time segment based on frequency response scheduling; the power system cost model includes: piecewise linear generation cost model, frequency response capacity cost model, frequency response deployment cost model, and wind energy curtailment penalty cost model;

[0022] The frequency response dispatch cost model is constructed to minimize the power system cost under the joint constraints of the power system.

[0023] Preferably, the piecewise linear power generation cost model satisfies the following relationship:

[0024]

[0025] In the formula, seg represents the time segment, a represents the generator, C sl represents the piecewise linear generation cost, represents the unit price of the frequency response capacity cost of generator a, represents the available up-regulated frequency response of generator a over time t, represents the available downregulated frequency response of generator a over time t, represents the linear cost unit price of generator a in time segment seg, It represents the actual power generation of generator a in time segment seg changing with time t, Δt is the frequency response scheduling duration, A is the number of generators, T is the duration of generator output adjustment, and N is the number of segments.

[0026] Preferably, the frequency response capacity cost model satisfies the following relationship:

[0027]

[0028] In the formula, C fr represents the frequency response capacity cost, λ t,s (t) is a Boolean variable indicating the occurrence of wind power generation scenario s at time t. When the value is 1, it indicates that wind power generation scenario s occurs at time t. represents the unit price of frequency response deployment cost of generator a, represents the upward frequency response of generator a deployment over time t in wind power generation scenario s, It represents the downward frequency response of generator a deployment changing with time t under wind power generation scenario s, and S is the number of wind power generation scenarios.

[0029] Preferably, the frequency response deployment cost model satisfies the following relationship:

[0030]

[0031] In the formula, C fr,d represents the frequency response deployment cost, represents the unit price of the flexible load reduction compensation cost of node e, It represents the flexible load reduction of node e over time t in wind power generation scenario s, and E is the number of nodes.

[0032] Preferably, the wind energy reduction penalty cost model satisfies the following relationship:

[0033]

[0034] Where P y,s Indicates wind energy reduction penalty costs, represents the unit price of wind curtailment compensation cost of wind farm y, It represents the wind power abandonment of wind farm y changing with time t under wind power generation scenario s, where Y is the number of wind farms.

[0035] Preferably, the joint constraints of the power system include: power generation constraints, power balance constraints at each node in the power system, thermal limit constraints of transmission lines, frequency response scheduling availability constraints for each generator, flexible load reduction constraints, wind power generation reduction constraints for each wind farm, 5-minute ramping constraints for each generator, flexible rising constraints and flexible falling constraints for each generator, flexible rising deployment constraints and flexible falling deployment constraints for each generator, and voltage angle constraints for reference nodes.

[0036] Preferably, based on the frequency response dispatch cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, the power generation cost difference corresponding to each time segment under the two wind power generation scenarios, the wind power reduction penalty cost difference and the frequency response deployment cost difference are determined respectively, including:

[0037] In the process of executing frequency response scheduling, based on the piecewise linear power generation cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty, the first power generation cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty is calculated; according to the operating data of the wind power generation scenario with wind speed uncertainty, the second power generation cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty is calculated; the difference between the first power generation cost and the second power generation cost in each time segment is used as the power generation cost difference corresponding to each time segment in the two wind power generation scenarios;

[0038] In the process of executing frequency response scheduling, based on the frequency response capacity cost model, Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty in each time segment are determined respectively;

[0039] Based on the frequency response deployment cost model, according to the Boolean variables of the wind power generation scenarios without wind speed uncertainty and the wind power generation scenarios with wind speed uncertainty appearing in each time segment, the first frequency response deployment cost corresponding to each time segment under the wind power generation scenario without wind speed uncertainty and the second frequency response deployment cost corresponding to each time segment under the wind power generation scenario with wind speed uncertainty are calculated respectively; the difference between the first frequency response deployment cost and the second frequency response deployment cost under each time segment is taken as the frequency response deployment cost difference corresponding to each time segment under the two wind power generation scenarios;

[0040] Based on the wind energy reduction penalty cost model, according to the Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty appearing in each time segment, the first wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty and the second wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty are calculated respectively; the difference between the first wind energy reduction penalty cost and the second wind energy reduction penalty cost in each time segment is taken as the difference in wind energy reduction penalty costs corresponding to each time segment in the two wind power generation scenarios.

[0041] The present invention also proposes a wind farm price stochastic optimization system based on wind speed uncertainty, comprising: a scenario generation module, a cost difference calculation module, and a wind farm price optimization module;

[0042] A scenario generation module is used to collect historical wind speed data of wind farms and generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty;

[0043] A cost difference calculation module is used to establish a frequency response dispatch cost model; based on the frequency response dispatch cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, the power generation cost difference, the frequency response deployment cost difference and the wind energy reduction penalty cost difference corresponding to each time segment under the two wind power generation scenarios are determined respectively;

[0044] The wind farm price optimization module is used to take the sum of the power generation cost difference, frequency response deployment cost difference and wind energy reduction penalty cost difference corresponding to each time segment under two wind power generation scenarios as the time-of-use price of the wind farm.

[0045] Preferably, the random optimization model establishing unit is used to establish a random optimization model, which takes the historical wind speed data of the wind farm as input and the wind power generation scenario as output; the random optimization model has multiple parallel scenario generation layers built in, and each scenario generation layer outputs a corresponding wind power generation scenario based on a different probability distribution model; the random optimization model is used to generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty.

[0046] Preferably, the cost difference calculation module includes: a power generation cost difference calculation unit, a scenario variable calculation unit, a frequency response deployment cost difference calculation unit, and a wind energy curtailment penalty cost difference calculation unit;

[0047] In the process of executing frequency response scheduling, based on the piecewise linear power generation cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty, the first power generation cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty is calculated; according to the operating data of the wind power generation scenario with wind speed uncertainty, the second power generation cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty is calculated; the difference between the first power generation cost and the second power generation cost in each time segment is used as the power generation cost difference corresponding to each time segment in the two wind power generation scenarios;

[0048] A scenario variable calculation unit, used to determine Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty in each time segment based on a frequency response capacity cost model during the frequency response scheduling process;

[0049] A frequency response deployment cost difference calculation unit is used to calculate, based on the frequency response deployment cost model, the first frequency response deployment cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty and the second frequency response deployment cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty according to the Boolean variables of the wind power generation scenario without wind speed uncertainty and the wind power generation scenario with wind speed uncertainty in each time segment; and take the difference between the first frequency response deployment cost and the second frequency response deployment cost in each time segment as the frequency response deployment cost difference corresponding to each time segment in the two wind power generation scenarios;

[0050] A wind energy reduction penalty cost difference calculation unit is used to calculate, based on a wind energy reduction penalty cost model and according to the Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty that appear in each time segment, respectively the first wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty and the second wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty; the difference between the first wind energy reduction penalty cost and the second wind energy reduction penalty cost in each time segment is used as the wind energy reduction penalty cost difference corresponding to each time segment in the two wind power generation scenarios.

[0051] A terminal comprises a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of a method.

[0052] A computer-readable storage medium stores a computer program, which implements the steps of the method when executed by a processor.

[0053] The beneficial effects of the present invention are that, compared with the prior art, at least the probabilities of occurrence of different scenarios in wind power generation are taken into consideration, and FR scheduling using only one uncertainty scenario is proposed, that is, a wind power generation scenario without wind speed uncertainty; FR scheduling using a large number of wind power generation scenarios is also proposed, that is, a wind power generation scenario with wind speed uncertainty, which not only reduces the prediction error and reduces the power dispatching cost, but also greatly reduces the calculation time for microgrid dispatching compared to a large number of scenarios, and is more real-time. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flow chart of a stochastic optimization method of wind farm price based on wind speed uncertainty proposed by the present invention;

[0055] Figure 2 is a wind speed scenario diagram obtained based on gamma distribution in an embodiment of the present invention;

[0056] Figure 3 This is the uncertainty price solution process diagram of this method. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.

[0058] The present invention proposes a stochastic optimization method for wind farm prices based on wind speed uncertainty, which is implemented based on the power system optimization model of FR in the stochastic optimization framework, such as Figure 1 As shown, the method includes:

[0059] Step 1: Collect historical wind speed data of the wind farm to generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty.

[0060] Specifically, a stochastic optimization model is established, which takes the historical wind speed data of the wind farm as input and the wind power generation scenario as output; the stochastic optimization model has multiple parallel scenario generation layers built in, and each scenario generation layer outputs the corresponding wind speed scenario based on different probability distribution models;

[0061] The stochastic optimization model is used to generate wind power generation scenarios without wind speed uncertainty and with wind speed uncertainty, including:

[0062] 1) Sample the collected historical wind speed data and verify the sampling values ​​using the time series sampling method;

[0063] Specifically, after obtaining the historical wind speed data of a certain place through website resources, the historical wind speed data of the place cannot be directly used for simulation, but it is necessary to sample the obtained historical wind speed data and verify the sampled wind speed data to obtain the representative wind speed distribution scenario of the place. Common sampling methods include time series sampling, random sampling, cluster center sampling, etc. And for the obtained sampled data, it is necessary to verify its validity, and the sampled data are generally processed by machine language or deep learning. Due to the large time scale, the wind speed distribution of a certain place has a certain regularity. The present invention uses the kaggle platform to obtain a wind speed data set of a certain place, and a part of the wind speed data set is used for sampling and a part for verification.

[0064] In a non-limiting preferred embodiment, the historical wind speed data obtained is sampled using a time series sampling method. The sampling time is one hour, and the sampled data is obtained. The sampled data is verified with a portion of the wind speed data of a certain place, and the time series verification method in deep learning is used. This verification method maintains the time order of the data, divides the data set into continuous time periods, and then cross-validates on each time period to obtain data for verifying the time series sampling.

[0065] 2) Calculate the mean and standard deviation of wind speed in each time interval based on the sampled and verified historical wind speed data; Generate corresponding wind speed scenarios using multiple scenario generation layers based on the mean and standard deviation of wind speed in each time interval.

[0066] In a non-limiting preferred embodiment, the probability distributions for generating multiple wind speed scenarios include, but are not limited to, gamma distribution, Weibull distribution, and Rayleigh distribution; those skilled in the art may select different probability distributions for each scenario generation layer according to actual needs.

[0067] In a non-limiting preferred embodiment, taking the standard form of gamma distribution as an example, the probability distribution model of different times of the day (preferably one hour) is set to be the same, and a representative wind speed distribution scenario of the area is generated based on the wind speed data obtained after sampling and verification processing. A matlab program is used to generate a wind speed scenario that obeys the gamma distribution, such as Figure 2 As shown. Therefore, the first wind speed scenario is generated based on the gamma distribution, the second wind speed scenario is generated based on the Weibull distribution, and the third wind speed scenario is generated based on the Rayleigh distribution. The Weibull distribution is not suitable for areas with very low or very high wind speeds. Meanwhile, the gamma distribution is another widely used distribution for wind speed modeling, and is suitable for areas with very low or very high wind speeds, and for areas with different underlying ground and climatic conditions. The Rayleigh distribution is a special form of the Weibull distribution. Since the parameters of the Rayleigh distribution are easy to estimate, it is also commonly used in wind speed modeling and wind turbine performance evaluation.

[0068] In addition, the wind speed mean and standard deviation of each time interval are calculated based on the historical wind speed data after sampling and verification, which can effectively improve the accuracy of the wind speed scene. In a non-limiting preferred embodiment, the parameter deviations of the gamma distribution before and after sampling and verification include: mean deviation Δμ is -0.0002, standard deviation deviation Δσ is 0.162, abundance deviation Δκ is 3.879, and skewness deviation Δγ is 0.7457. It can be seen from the parameter deviation that the present invention is based on a random optimization model, which can not only output wind speed scenes that obey multiple probability distribution models at the same time, but also improve the accuracy of the wind speed scene. After processing with the random optimization model, a representative wind speed distribution scene of a certain place can be obtained.

[0069] 3) A single wind speed scenario is used as a wind power generation scenario without wind speed uncertainty, and multiple wind speed scenarios are used to form a wind power generation scenario with wind speed uncertainty;

[0070] However, in practical applications, if the uncertainty of wind speed is taken into account, the probability density function of the prediction error at each moment is different. Therefore, different probability distribution models should be established in different time periods to obtain the probability of occurrence of each wind speed scenario in a year and the probability of occurrence in each time period of a day; in a non-restrictive preferred embodiment, the single wind speed scenario with the highest probability of occurrence in a year is used as the wind power generation scenario without wind speed uncertainty; or, a single wind speed scenario generated by the same probability distribution model is used for all time periods in a day as the scenario without wind speed uncertainty; the wind speed scenario with the highest probability of occurrence in each time period of a day is extracted, and the extracted wind speed scenarios are spliced ​​in the order of time periods to obtain a wind power generation scenario with wind speed uncertainty.

[0071] In the prior art, when the same wind speed scenario generated by probability distribution is used for scheduling calculation of integrated energy microgrid in all periods of a day, the difference between prediction errors at different times caused by wind speed uncertainty is ignored, resulting in inaccurate calculation in certain periods; however, if the scheduling calculation of integrated energy microgrid is performed according to a large number of different wind speed scenarios, the disadvantage of inaccurate calculation in certain periods is alleviated to a certain extent through the mutual complementation of multiple scenarios, so that the accuracy of the calculation results is significantly improved, but the calculation amount is huge and the calculation time is very long; moreover, before using a large number of different wind speed scenarios for scheduling calculation of integrated energy microgrid, it is necessary to screen the wind speed scenarios based on the probability of occurrence of the scenarios. In the screening process, the influence of some extreme scenarios and the influence of some scenarios related to periodic climate are ignored, thereby affecting the scope of application of the calculation results. The present invention is directed to the above-mentioned problem and proposes that: the wind speed scenarios generated by the same probability distribution in all periods of a day can be used as scenarios not based on wind speed uncertainty, and multiple wind speed scenarios generated by different probability distributions in different periods of a day are spliced ​​according to the time series relationship to obtain scenarios based on wind speed uncertainty, which not only retains the difference between the prediction errors at different times caused by wind speed uncertainty, but also effectively reduces the number of calculation scenarios, and reduces the amount of calculation and shortens the calculation time while ensuring the accuracy of the calculation results. Moreover, when the wind farm is affected by periodic climate, the probability of occurrence of some scenarios in fixed periods of a day increases sharply. The method proposed by the present invention effectively takes these specific scenarios into account, thereby improving the reliability of scheduling. Existing studies have shown that the impact of climate change on wind resources has changed the temporal and spatial distribution of wind speed scenarios, resulting in the emergence of certain extreme wind speed scenarios; affecting the interannual wind power in the region; changing the wind energy potential in the region, etc. Therefore, the present invention has foresight and certain adaptability in responding to the change of wind speed scenarios in response to climate change. The wind speed uncertainty scenario obtained by the present invention can provide a basis for wind power pricing in the context of climate change.

[0072] Furthermore, step 1 also includes: calculating operating data in a wind power generation scenario without wind speed uncertainty, and calculating operating data in a wind power generation scenario with wind speed uncertainty, wherein the operating data in each wind power generation scenario includes but is not limited to: wind power generation.

[0073] The wind power generation model considers three important values: the cut-in wind speed (the wind speed below which the wind turbine power output is zero), the cut-out wind speed (the wind speed above which the wind turbine power output is zero), and the rated wind speed of the turbine. When the wind speed is between the rated wind speed and the cut-off wind speed, the wind turbine outputs the rated power. Therefore, the wind power generation is calculated based on the wind speed using the following relationship:

[0074]

[0075] Where P r,s Represents wind power generation, P rate Indicates the rated power output of the wind turbine, ν ci represents the cut-in wind speed, ν co represents the cut-out wind speed, ν rate Indicates rated wind speed.

[0076] Step 2: Establish a frequency response scheduling cost model.

[0077] Specifically, a day is divided into multiple time segments, and a power system cost model for each time segment is established based on frequency response scheduling, wherein the power system cost model includes: a piecewise linear power generation cost model, a frequency response capacity cost model, a frequency response deployment cost model, and a wind energy reduction penalty cost model; the frequency response scheduling cost model is constructed to minimize the power system cost under the joint constraints of the power system.

[0078] Specifically, step 2 includes:

[0079] Step 2.1: Divide a day into multiple time segments, and establish a power system cost model for each time segment based on frequency response scheduling, including:

[0080] 1) In any time segment, the relevant economic indicators when the generator adjusts its output according to the frequency response dispatch instruction are used as the piecewise linear power generation cost. The piecewise linear power generation cost model satisfies the following relationship:

[0081]

[0082] In the formula, seg represents the time segment, a represents the generator, and C sl represents the piecewise linear generation cost, represents the unit price of the frequency response capacity cost of generator a, represents the available up-regulated frequency response of generator a over time t, represents the available downregulated frequency response of generator a over time t, represents the linear cost unit price of generator a in time segment seg, It represents the actual power generation of generator a in time segment seg changing with time t, Δt is the frequency response scheduling duration, A is the number of generators, T is the duration of generator output adjustment, and N is the number of segments.

[0083] The available up-regulation frequency response and the available down-regulation frequency response of each wind turbine are presented as a function of time, making the piecewise linear generation cost more accurate.

[0084] The cost calculation of the power system is divided into segments, and the upward and downward frequency responses of the generator are taken as a function of time. Even if the volatility and uncertainty of wind speed will have an impact on the power system and cause system frequency instability, the required piecewise linear power generation cost can be accurately calculated, reducing the error and the dispatch cost of the power grid, while improving the reliability of the power grid cost calculation.

[0085] 2) In the wind power generation scenario, the relevant economic indicators when the generator adjusts the frequency according to the frequency response dispatch instruction are used as the frequency response capacity cost. The frequency response capacity cost model satisfies the following relationship:

[0086]

[0087] In the formula, C fr represents the frequency response capacity cost, λ t,s (t) is a Boolean variable indicating the occurrence of wind power generation scenario s at time t. When the value is 1, it indicates that wind power generation scenario s occurs at time t. represents the unit price of frequency response deployment cost of generator a, represents the upward frequency response of generator a deployment over time t in wind power generation scenario s, It represents the downward frequency response of generator a deployment changing with time t under wind power generation scenario s, and S is the number of wind power generation scenarios.

[0088] The upward frequency response and downward frequency response of each wind turbine in the wind power generation scenario are taken as a function that varies with time, so that the frequency response capacity cost in the wind power generation scenario is more accurate.

[0089] In the formula, the frequency response capacity cost under different wind power generation scenarios is taken into account. The existing methods generally only use one wind power generation scenario to calculate the frequency response capacity cost. Compared with the existing methods, the frequency response capacity cost calculated using this formula can adapt to a wider range of wind power generation scenarios, and the calculated frequency response capacity cost under complex wind power generation scenarios is more accurate.

[0090] 3) In the wind power generation scenario, the relevant economic indicators when the flexible load adjusts the capacity according to the frequency response dispatch instruction are used as the frequency response deployment cost. The frequency response deployment cost model satisfies the following relationship:

[0091]

[0092] In the formula, Cf r,d represents the frequency response deployment cost, represents the unit price of the flexible load reduction compensation cost of node e, It represents the flexible load reduction of node e over time t in wind power generation scenario s, and E is the number of nodes.

[0093] In the formula, the frequency response deployment cost under different wind power generation scenarios is taken into account. The existing methods generally only use one wind power generation scenario to calculate the frequency response deployment cost. Compared with the existing methods, the frequency response deployment cost calculated using this formula can adapt to more complex wind power generation scenarios. The frequency response deployment cost calculated under different wind power generation scenarios has better adaptability and is more accurate.

[0094] 4) In the wind power generation scenario, according to the frequency response dispatch instruction, the wind energy reduction penalty cost model satisfies the following relationship:

[0095]

[0096] Where P y,s Indicates wind energy reduction penalty costs, represents the unit price of wind curtailment compensation cost of wind farm y, It represents the wind power abandonment of wind farm y changing with time t under wind power generation scenario s, where Y is the number of wind farms.

[0097] Step 2.2, establish the joint constraints of the power system.

[0098] Specifically, the joint constraints of the power system include:

[0099] 1) The power generation constraints satisfy the following relationship:

[0100]

[0101] Where Pat represents the actual power generation of generator a at time t, represents the actual power generation of generator a in the segment seg at time t, Indicates the upper limit of power generation of generator a in segment seg.

[0102] 2) The power balance constraint of each node in the power system satisfies the following relationship:

[0103]

[0104] In the formula, NG b NW represents the set of generators connected to node n, i represents the generator in the set of generators; b represents the set of wind farms connected to node n, and w represents the wind farms in the set of wind farms; represents the set of transmission lines connecting the “destination” bus to node n, represents the set of transmission lines connecting the “starting point” bus and the node n, and l represents the transmission line in the set of transmission lines; represents the upward frequency response of generator a deployment at time t under scenario s, represents the down-regulated frequency response of generator a deployed at time t under scenario s, represents the wind power generation of wind farm y at time t under scenario s, represents the amount of wind power abandoned by wind farm y at time t under scenario s, PL lt,s represents the actual power flow through transmission line l at time t under scenario s, represents the load of node c at time t, represents the flexible load reduction amount of node c at time t under scenario s.

[0105] 3) The thermal constraint of the transmission line satisfies the following relationship:

[0106]

[0107] In the formula, represents the maximum practical power flow limit of transmission line l, P lt,s represents the actual power flow through the transmission line l at time t under scenario s.

[0108] 4) The frequency response dispatch availability constraint of each generator satisfies the following relationship:

[0109]

[0110] Where P at represents the actual power generation of generator a at time t, represents the upper limit of power generation of generator a, represents the lower limit of power generation of generator a, represents the available upward frequency response setpoint of generator a, Represents the available downregulated frequency response setpoint for generator a.

[0111] 5) The flexible load reduction constraint condition satisfies the following relationship:

[0112]

[0113] In the formula, represents the flexible load reduction of node e at time t under scenario s, represents the load of node e at time t.

[0114] 6) The wind power reduction constraint of each wind farm satisfies the following relationship:

[0115]

[0116] In the formula, represents the amount of wind power abandoned by wind farm y at time t under scenario s, represents the wind power generation of wind farm y at time t under scenario s.

[0117] 7) The 5-minute ramping constraint condition of each generator satisfies the following relationship:

[0118]

[0119] In the formula, RU a Indicates the power increase rate of generator a per minute, RD a Indicates the power reduction rate of generator a per minute.

[0120] 8) The flexible rising constraint condition and flexible falling constraint condition of each generator satisfy the following relationship respectively:

[0121]

[0122] 9) The flexible rising deployment constraint and the flexible falling deployment constraint of each generator satisfy the following relations:

[0123]

[0124] In the formula, represents the upward frequency response of generator a deployment at time t under scenario s, represents the downward frequency response of generator a deployment at time t under scenario s.

[0125] 10) Set node 1 as the reference node, and the voltage angle constraint of the reference node satisfies the following relationship:

[0126]

[0127] In the formula, β 1,t,s represents the voltage angle of the reference node at time t under scenario s, and the voltage angle of different generators is different.

[0128] In step 2.3, the frequency response dispatch cost model is constructed by minimizing the power system cost under the joint constraints of the power system.

[0129] When establishing the power system cost model, the present invention characterizes the impact of wind power generation uncertainty on the power system through frequency response capacity cost and frequency response deployment cost, and in different time periods, the minimum power system cost min (C sl +C fr +C fr,d +P y,s ) is the optimal value of the frequency response scheduling cost, thereby achieving the optimal frequency response scheduling.

[0130] Step 3, based on the frequency response dispatch cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, determine the power generation cost difference, wind power reduction penalty cost difference and frequency response deployment cost difference corresponding to each time segment under the two wind power generation scenarios.

[0131] In the process of executing frequency response scheduling, the present invention calculates the power generation cost difference, wind power curtailment compensation difference, and flexible load curtailment penalty difference according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, respectively, and uses various differences under the two scenarios to characterize the impact of wind speed uncertainty on the wind farm price, which can more accurately and representatively characterize the response to frequency response scheduling.

[0132] Specifically, step 3 includes:

[0133] Step 3.1, in the process of executing frequency response scheduling, based on the piecewise linear power generation cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty, calculate the first power generation cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty, and according to the operating data of the wind power generation scenario with wind speed uncertainty, calculate the second power generation cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty; the difference between the first power generation cost and the second power generation cost in each time segment is used as the power generation cost difference corresponding to each time segment in the two wind power generation scenarios;

[0134] Step 3.2, in the process of executing the frequency response dispatch, based on the frequency response capacity cost model, respectively determine the Boolean variables of the wind power generation scenarios without wind speed uncertainty and the wind power generation scenarios with wind speed uncertainty in each time segment;

[0135] Step 3.3, based on the frequency response deployment cost model, according to the Boolean variables of the wind power generation scenarios without wind speed uncertainty and the wind power generation scenarios with wind speed uncertainty in each time segment, respectively calculate the first frequency response deployment cost corresponding to each time segment under the wind power generation scenario without wind speed uncertainty and the second frequency response deployment cost corresponding to each time segment under the wind power generation scenario with wind speed uncertainty; the difference between the first frequency response deployment cost and the second frequency response deployment cost in each time segment is taken as the frequency response deployment cost difference corresponding to each time segment under the two wind power generation scenarios;

[0136] Step 3.4, based on the wind energy reduction penalty cost model, according to the Boolean variables of the wind power generation scenarios without wind speed uncertainty and the wind power generation scenarios with wind speed uncertainty that appear in each time segment, respectively calculate the first wind energy reduction penalty cost corresponding to each time segment under the wind power generation scenario without wind speed uncertainty and the second wind energy reduction penalty cost corresponding to each time segment under the wind power generation scenario with wind speed uncertainty; the difference between the first wind energy reduction penalty cost and the second wind energy reduction penalty cost in each time segment is taken as the difference in wind energy reduction penalty costs corresponding to each time segment under the two wind power generation scenarios.

[0137] The present invention improves the frequency response cost of wind power generation, improves the method of calculating the frequency response cost, and regards the frequency response as a function that changes over time. Compared with the prior art, it has the advantages of more accurate calculation and reduced grid dispatching cost. In the frequency response dispatching (FR) process, the scenario of frequency response dispatching is improved. When the prior art adopts frequency response dispatching, only one uncertain scenario is used for frequency response dispatching, and the error is large. The present invention takes into account the impact of multiple uncertain scenarios on the frequency response dispatching cost, and can obtain the frequency response dispatching cost more accurately and representatively.

[0138] Step 4: The sum of the power generation cost difference, frequency response deployment cost difference and wind energy reduction penalty cost difference corresponding to each time segment under the two wind power generation scenarios is used as the time-of-use price of the wind farm.

[0139] In an embodiment of the present invention, to determine the influence of wind speed uncertainty, three probability functions are used to create 200 wind speed scenarios for wind speed distribution, with a resolution of 5 minutes and a duration of one day. To generate these scenarios, wind speed data of a certain place is used, and the wind speed mean and standard deviation of each 5-minute time point in January are calculated. 288 shape parameters c and scale parameters p (with 24 hours and every 5-minute interval as the time point) are created using the wind speed mean and standard deviation at each time point, and 288 different error distribution parameters are obtained. Using these distribution parameters, 200 random wind speed values ​​are generated at each time point.

[0140] Using the generated wind speed scenarios, the wind turbine output power was calculated for each wind speed. The model used a cut-in speed of 4 m / s, a cut-out speed of 25 m / s, a rated wind speed of 14 m / s, and a rated output power of 200 MW for each wind farm. Using 200 wind speed values ​​for each 5-minute interval, 200 wind output scenarios for each interval of each day were created for each probability function (PDF).

[0141] like Figure 3 As shown, the method is investigated in five cases:

[0142] 1) The stochastic optimization FR scheduling problem only considers one case and assumes that the wind power generation is always zero in this case.

[0143] 2) Consider only one case and use the average wind speed at each time point to generate the wind output case.

[0144] 3) 200 scenarios generated by gamma distribution were considered.

[0145] 4) 200 scenarios generated by Rayleigh distribution were considered.

[0146] 5) 200 scenarios generated by Weibull distribution were considered.

[0147] In the stochastic optimization dispatch model, wind power curtailment and flexible load curtailment are allowed, but they are penalized in the objective function. The penalties for wind power curtailment and flexible load curtailment are 210 yuan per MW and 60,000 yuan per MW, respectively. Since there is no uncertainty in both cases, curtailment is not allowed.

[0148] Formula (4) is used to calculate the wind power reduction cost to be 210 yuan per megawatt. Flexible load refers to a load that actively participates in the operation and control of the power grid, can interact with the power grid, and has flexible characteristics. The flexible load cost is obtained through the regulation of the power grid, and in this invention, it is taken as 60,000 yuan per megawatt.

[0149] The volatility and uncertainty of wind speed will impact the power system and cause the system frequency to be unstable. Therefore, we segment the cost calculation of the power system and regard the up-and-down frequency response of the generator as a time-varying function. In this way, the required piecewise linear power generation cost can be calculated more accurately, reducing the error and the dispatching cost of the power grid, and improving the reliability of the power grid cost calculation. Accurate wind power forecasting is crucial to the economic dispatch of the power system. The forecast error will affect the dispatching decision of the power system. Therefore, we consider the probability of different scenarios in wind power generation and propose FR dispatching using only one uncertainty scenario, that is, no uncertainty; and FR dispatching using a large number of wind power generation scenarios, both uncertainties. A comparison was made and it was found that FR dispatching using only one uncertainty scenario is not accurate; however, the calculation of integrated energy microgrid dispatching based on a large number of scenarios will have a very large amount of calculation, so we selected several representative scenarios and their probabilities according to demand, which not only reduces the forecast error and reduces the power dispatching cost, but also greatly reduces the calculation time of microgrid dispatching compared with a large number of scenarios, and is more real-time. The frequency response cost of wind power generation is improved, and the method for calculating the frequency response cost is improved. The frequency response is regarded as a function that changes over time. Compared with the prior art, it has the advantages of more accurate calculation and reduced grid dispatching cost. In the frequency response dispatching (FR) process, the scenario of frequency response dispatching is improved. When the prior art adopts frequency response dispatching, only one uncertain scenario is used for frequency response dispatching, and the error is large. The present invention takes into account the impact of multiple uncertain scenarios on the frequency response dispatching cost, and can obtain the frequency response dispatching cost more accurately and representatively.

[0150] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0151] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0152] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0153] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0154] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A stochastic optimization method for wind farm prices based on wind speed uncertainty, characterized in that: include: Collect historical wind speed data of wind farms to generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty; Establish a frequency response dispatch cost model; Based on the frequency response dispatch cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, the power generation cost difference, frequency response deployment cost difference and wind energy curtailment penalty cost difference corresponding to each time segment under the two wind power generation scenarios are determined respectively; The sum of the power generation cost difference, frequency response deployment cost difference and wind energy reduction penalty cost difference corresponding to each time segment under the two wind power generation scenarios is taken as the time-of-use price of the wind farm.

2. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 1 is characterized in that: Establish a stochastic optimization model that takes historical wind speed data of wind farms as input and wind power generation scenarios as output; The stochastic optimization model has multiple parallel scenario generation layers built in, and each scenario generation layer outputs the corresponding wind power generation scenario based on different probability distribution models; The stochastic optimization model is used to generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty.

3. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 2 is characterized in that: The stochastic optimization model is used to generate wind power generation scenarios without wind speed uncertainty and with wind speed uncertainty, including: The collected historical wind speed data are sampled and the sampling values ​​are verified using the time series sampling method; The wind speed mean and standard deviation of each time interval are calculated based on the sampled and verified historical wind speed data; the corresponding wind power generation scenarios are generated using multiple scenario generation layers based on the wind speed mean and standard deviation of each time interval; A single wind speed scenario is used as a wind power generation scenario without wind speed uncertainty, and multiple wind speed scenarios constitute a wind power generation scenario with wind speed uncertainty.

4. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 1, characterized in that: Establish a frequency response dispatch cost model, including: Divide a day into multiple time segments, and establish a power system cost model for each time segment based on frequency response scheduling; the power system cost model includes: piecewise linear generation cost model, frequency response capacity cost model, frequency response deployment cost model, and wind energy curtailment penalty cost model; The frequency response dispatch cost model is constructed to minimize the power system cost under the joint constraints of the power system.

5. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 4 is characterized in that: The piecewise linear power generation cost model satisfies the following relationship: In the formula, seg represents the time segment, a represents the generator, and C sl represents the piecewise linear generation cost, represents the unit price of the frequency response capacity cost of generator a, represents the available up-regulated frequency response of generator a over time t, represents the available downregulated frequency response of generator a over time t, represents the linear cost unit price of generator a in time segment seg, It represents the actual power generation of generator a in time segment seg changing with time t, Δt is the frequency response scheduling duration, A is the number of generators, T is the duration of generator output adjustment, and N is the number of segments.

6. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 5 is characterized in that: The frequency response capacity cost model satisfies the following relationship: In the formula, C fr represents the frequency response capacity cost, λ t,s (t) is a Boolean variable indicating the occurrence of wind power generation scenario s at time t. When the value is 1, it indicates that wind power generation scenario s occurs at time t. represents the unit price of frequency response deployment cost of generator a, represents the upward frequency response of generator a deployment over time t in wind power generation scenario s, It represents the downward frequency response of generator a deployment changing with time t under wind power generation scenario s, and S is the number of wind power generation scenarios.

7. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 6 is characterized in that: The frequency response deployment cost model satisfies the following relationship: In the formula, C fr,d represents the frequency response deployment cost, represents the unit price of the flexible load reduction compensation cost of node e, It represents the flexible load reduction of node e over time t in wind power generation scenario s, and E is the number of nodes.

8. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 7 is characterized in that: The wind energy reduction penalty cost model satisfies the following relationship: Where P y,s Indicates wind energy reduction penalty costs, represents the unit price of wind curtailment compensation cost of wind farm y, It represents the wind power abandonment of wind farm y changing with time t under wind power generation scenario s, where Y is the number of wind farms.

9. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 8, characterized in that: The joint constraints of the power system include: power generation constraints, power balance constraints at each node in the power system, thermal limit constraints of transmission lines, frequency response dispatch availability constraints for each generator, flexible load reduction constraints, wind power generation reduction constraints for each wind farm, 5-minute ramping constraints for each generator, flexible ramp-up constraints and flexible ramp-down constraints for each generator, flexible ramp-up deployment constraints and flexible ramp-down deployment constraints for each generator, and voltage angle constraints for reference nodes.

10. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 9, characterized in that: Based on the frequency response dispatch cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, the power generation cost difference, wind power reduction penalty cost difference and frequency response deployment cost difference corresponding to each time segment under the two wind power generation scenarios are determined respectively, including: In the process of executing frequency response scheduling, based on the piecewise linear power generation cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty, the first power generation cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty is calculated; according to the operating data of the wind power generation scenario with wind speed uncertainty, the second power generation cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty is calculated; the difference between the first power generation cost and the second power generation cost in each time segment is used as the power generation cost difference corresponding to each time segment in the two wind power generation scenarios; In the process of executing frequency response scheduling, based on the frequency response capacity cost model, Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty in each time segment are determined respectively; Based on the frequency response deployment cost model, according to the Boolean variables of the wind power generation scenarios without wind speed uncertainty and the wind power generation scenarios with wind speed uncertainty appearing in each time segment, the first frequency response deployment cost corresponding to each time segment under the wind power generation scenario without wind speed uncertainty and the second frequency response deployment cost corresponding to each time segment under the wind power generation scenario with wind speed uncertainty are calculated respectively; the difference between the first frequency response deployment cost and the second frequency response deployment cost under each time segment is taken as the frequency response deployment cost difference corresponding to each time segment under the two wind power generation scenarios; Based on the wind energy reduction penalty cost model, according to the Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty appearing in each time segment, the first wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty and the second wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty are calculated respectively; the difference between the first wind energy reduction penalty cost and the second wind energy reduction penalty cost in each time segment is taken as the difference in wind energy reduction penalty costs corresponding to each time segment in the two wind power generation scenarios.

11. A wind farm price stochastic optimization system based on wind speed uncertainty, characterized in that: include: Scenario generation module, cost difference calculation module, wind farm price optimization module; A scenario generation module is used to collect historical wind speed data of wind farms and generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty; A cost difference calculation module is used to establish a frequency response dispatch cost model; based on the frequency response dispatch cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty and the operating data of the wind power generation scenario with wind speed uncertainty, the power generation cost difference, the frequency response deployment cost difference and the wind energy reduction penalty cost difference corresponding to each time segment under the two wind power generation scenarios are determined respectively; The wind farm price optimization module is used to take the sum of the power generation cost difference, frequency response deployment cost difference and wind energy reduction penalty cost difference corresponding to each time segment under two wind power generation scenarios as the time-of-use price of the wind farm.

12. The wind farm price stochastic optimization system based on wind speed uncertainty according to claim 11, characterized in that: The scenario generation module includes: random optimization model building unit, The random optimization model establishment unit is used to establish a random optimization model. The random optimization model takes the historical wind speed data of the wind farm as input and takes the wind power generation scenario as output. The random optimization model has multiple parallel scenario generation layers built in, and each scenario generation layer outputs a corresponding wind power generation scenario based on a different probability distribution model. The random optimization model is used to generate wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty.

13. The method for stochastic optimization of wind farm prices based on wind speed uncertainty according to claim 11, characterized in that: The cost difference calculation module includes: a power generation cost difference calculation unit, a scenario variable calculation unit, a frequency response deployment cost difference calculation unit, and a wind energy reduction penalty cost difference calculation unit; In the process of executing frequency response scheduling, based on the piecewise linear power generation cost model, according to the operating data of the wind power generation scenario without wind speed uncertainty, the first power generation cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty is calculated; according to the operating data of the wind power generation scenario with wind speed uncertainty, the second power generation cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty is calculated; the difference between the first power generation cost and the second power generation cost in each time segment is used as the power generation cost difference corresponding to each time segment in the two wind power generation scenarios; A scenario variable calculation unit, used to determine Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty in each time segment based on a frequency response capacity cost model during the frequency response scheduling process; A frequency response deployment cost difference calculation unit is used to calculate, based on the frequency response deployment cost model, the first frequency response deployment cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty and the second frequency response deployment cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty according to the Boolean variables of the wind power generation scenario without wind speed uncertainty and the wind power generation scenario with wind speed uncertainty in each time segment; and take the difference between the first frequency response deployment cost and the second frequency response deployment cost in each time segment as the frequency response deployment cost difference corresponding to each time segment in the two wind power generation scenarios; A wind energy reduction penalty cost difference calculation unit is used to calculate, based on a wind energy reduction penalty cost model and according to the Boolean variables of wind power generation scenarios without wind speed uncertainty and wind power generation scenarios with wind speed uncertainty that appear in each time segment, respectively the first wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario without wind speed uncertainty and the second wind energy reduction penalty cost corresponding to each time segment in the wind power generation scenario with wind speed uncertainty; the difference between the first wind energy reduction penalty cost and the second wind energy reduction penalty cost in each time segment is used as the wind energy reduction penalty cost difference corresponding to each time segment in the two wind power generation scenarios.

14. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1 to 10.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.