A primary frequency regulation capacity planning method considering the joint participation of wind power clusters and shared energy storage

Abstract

The present invention relates to a planning model and solution method for wind power clusters to jointly share energy storage to participate in primary frequency regulation, specifically a method for planning primary frequency regulation capacity taking into account the joint participation of wind power clusters with shared energy storage. By establishing a wind power output uncertainty model based on the probability distribution of Weibull distribution, comprehensively considering the output characteristics of each wind turbine in the wind power cluster, centrally configuring shared energy storage to jointly undertake the primary frequency regulation task, and processing the relationship between reliability and economy through chance-constrained planning, a capacity optimization model for wind power clusters to jointly participate in the primary frequency regulation of the power grid with shared energy storage is established. Finally, a solution method for converting chance constraints into deterministic constraints is used to improve the efficiency of model solution. The scheme determined by the present invention realizes the coordinated planning of the economy and reliability of the primary frequency regulation of the wind-storage joint system when the number of wind turbines increases and the data of the wind power cluster is huge and complex, and has better solution efficiency and has a prospect for promotion and application.

Description

Technical Field

[0001] The present invention relates to planning for primary frequency regulation of renewable energy sources in power systems, and in particular to a capacity planning model and corresponding solution method for primary frequency regulation of wind power systems that takes into account wind power uncertainty, wind power clusters, shared energy storage, and opportunity-constrained planning. Specifically, the present invention relates to a capacity planning method for primary frequency regulation that takes into account the joint participation of wind power clusters and shared energy storage. Background Art

[0002] As a major source of renewable energy, wind power is poised to continue to increase in penetration within the power system, posing significant challenges to the safe and stable operation of traditional power systems. First, the randomness and volatility of wind speed—the amplitude, frequency, and distribution of its output fluctuations—are weakly correlated with grid load fluctuations, negatively impacting power system frequency stability. Second, the large-scale integration of wind power will displace traditional generators providing spinning reserve, reducing the grid's primary frequency regulation capabilities. To ensure safe and stable system operation, wind farms must possess primary frequency regulation capabilities.

[0003] Technologies for achieving primary frequency regulation in wind turbines include power reserve control and rotor kinetic energy control within the wind turbine itself. However, these technologies present issues such as limited rotor kinetic energy, secondary frequency drops, and prolonged wind curtailment that impacts profitability. Furthermore, wind turbines can achieve better primary frequency regulation by configuring appropriate amounts of energy storage, leveraging its fast response, flexible, and controllable characteristics. Capacity planning for the combined participation of wind and energy storage in grid primary frequency regulation essentially involves rationally balancing the cost of wind turbine curtailment and energy storage costs, while ensuring safe and reliable system operation, taking into account wind power uncertainty and related constraints, in order to achieve optimal reliability and economic efficiency for the power system.

[0004] Currently, wind-storage frequency regulation planning typically considers the combination of a single wind turbine and a single energy storage unit. However, with the increasing number of grid-connected wind turbines, the simultaneous configuration of energy storage in wind farms alone may result in insufficient energy storage utilization. It is necessary to transform the planning problem of single wind turbines combined with energy storage for primary frequency regulation into a planning problem of wind-storage for primary frequency regulation that takes into account the effects of wind farm clusters. Due to the wake effect, delay effect, and wind farm cluster effects influenced by topography, the wind speeds of each wind turbine at the same time vary, resulting in different operating states and, consequently, different frequency regulation capabilities for each wind turbine. If energy storage is configured for each wind turbine to meet uniform primary frequency regulation standards, the required energy storage capacity would be large, resulting in a certain degree of resource waste. Therefore, it is of great significance to study the centralized configuration of shared energy storage with a reasonable capacity for wind farm clusters to achieve wind farm resource complementarity, improve energy storage utilization efficiency, thereby reducing the required energy storage capacity and achieving better reliability and economic efficiency for wind-storage combined systems.

[0005] With the increasing number of wind turbines and the large and complex data volumes of wind power clusters, the currently used intelligent algorithms based on stochastic simulation have the disadvantages of requiring long computation times and being easily trapped in local optima when solving chance-constrained programming. This study transforms chance-constrained programming into deterministic constraints to achieve a more efficient and accurate solution that comprehensively considers factors such as wind power uncertainty and wind power cluster effects, thereby achieving a better planning method for the combined participation of wind and energy storage in primary frequency regulation. Summary of the Invention

[0006] In order to address the shortcomings of the existing technology in the process of studying the primary frequency regulation planning of the wind-storage joint participation system, the present invention proposes a primary frequency regulation planning method that takes into account the wind power cluster shared energy storage joint participation system. The scenario in which the present invention is applied is a wind power cluster composed of multiple wind turbines or wind farms with a close electrical distance. Due to the wind power wake effect between different wind farms, the wind speed and the frequency regulation capability that can be provided by each wind turbine or wind farm at the same time are different. Therefore, shared energy storage is centrally configured to achieve better reliability and economy. Taking into account the wind power cluster effect, multiple wind farms and shared energy storage are connected to the large power grid through the same grid connection point to jointly undertake the primary frequency regulation task.

[0007] The present invention is implemented by adopting the following technical solution: a method for primary frequency regulation capacity planning taking into account the joint participation of wind power cluster shared energy storage, comprising the following steps:

[0008] Step 1, fitting and generating wind speed data describing wind speed uncertainty;

[0009] Step 2: Consider the wind power cluster effect including wake effect, delay effect, and topography to establish a wind power cluster output uncertainty model;

[0010] Step 3: Considering the characteristics of multiple wind farms in a wind power cluster scenario, a capacity planning model for wind power clusters sharing energy storage and jointly participating in system primary frequency regulation based on chance-constrained programming is established.

[0011] Step 4: With a given confidence level, the convolution is quickly calculated using the Fast Fourier Transform (FFT) to obtain the probability distribution of the joint variables, formally transforming the chance constraint into a deterministic constraint, and then solving it to obtain the optimal capacity configuration plan for the wind power cluster load reduction rate and shared energy storage.

[0012] Furthermore, step 1 is specifically as follows: select the two-parameter Weibull distribution to fit the wind speed, and the distribution function is: Its probability density function can be expressed as: Its mean for: Its variance σ(v 0,t )for: Where: v 0,tTo fit the generated natural wind speed, k and c are the two parameters of the Weibull distribution, k is the shape parameter, c is the scale parameter, and Γ(·) is the gamma function.

[0013] Furthermore, step 2 is specifically as follows: the wind power cluster effect includes the wake effect of the wind turbine, the delay effect of the wind, and the influence of the topography where the wind farm is located;

[0014] The simulation of the wake effect in wind power cluster effect is based on the Jensen wake model. When the incoming wind speed is v 0,t , the wind speed at the hub height at a distance l downstream of the wind turbine is v t , the radius of the wind turbine impeller is r0, the wake radius at the downstream position l of the wind turbine is r=r0+sl, s is the velocity recovery coefficient, then the wind speed v of the downstream wind turbine affected by the upstream single wind turbine is t for: Where: C T is the thrust coefficient, s is the speed recovery coefficient;

[0015] There are many wind turbines in a wind farm cluster. The downstream wind turbines are located in the superposition area of the wakes of multiple upstream wind turbines. When downstream wind turbine i is affected by n upstream wind turbines, the kinetic energy loss of the mixed wakes of multiple wind turbines is equal to the sum of the kinetic energy losses of the wakes of a single wind turbine. Therefore, the actual wind speed of wind turbine i is: Where: v i1,t is the wind speed of wind turbine i after considering the wake effect, l ij is the radial distance between wind turbine i and wind turbine j;

[0016] The influence of the topography of the wind farm in the wind power cluster effect is simulated as follows. For a wind farm with undulating terrain, if A is the coverage area of the upstream wind turbine wake at the downstream wind turbine location, B is the rotation plane of the downstream wind turbine rotor, and the overlapping part is C, then the wake shielding area is used as the correction weight. After considering the influence of the wind power cluster's topography, the wind speed v of wind turbine i is i2,t for:

[0017] The delay effect in the wind power cluster effect is simulated by the delay time of different wind turbines. The wind turbines in the wind farm are a certain distance apart. It takes a certain delay time for the wind to blow from the first row of wind turbines to the second row of wind turbines. The delay time tl of the wind between the second row of wind turbines can be expressed as: v i3,t =v i2,t-tl , where L is the distance between two adjacent rows of fans along the wind speed direction, v i3,t The wind speed v for the wind power cluster taking into account the wake effect and the influence of topography i2,t The actual wind speed of wind turbine i after the delay time is applied;

[0018] The final wind speed-wind power relationship formula is as follows:

[0019]

[0020] Where: v in is the cut-in wind speed; .v out . is the cut-out wind speed; v r is the rated wind speed; P WNi is the rated capacity of the i-th wind turbine; P Wi,t is the output power of the i-th wind turbine under different wind speeds, thereby obtaining the wind power cluster output power considering wind speed uncertainty and wind power cluster effect.

[0021] Furthermore, step 3 is specifically as follows: in the planning of the primary frequency regulation capacity of the wind power cluster and the shared energy storage, the wind power cluster adopts wind power variable pitch control load shedding operation to realize the shared primary frequency regulation task of the system together with the shared energy storage frequency regulation. The primary frequency regulation reserve capacity that the i-th wind turbine in the wind power cluster can provide is: P Wri,t =P Wi,t x, where x is the load reduction percentage of the wind power cluster and the equivalent regulation coefficient of the wind power cluster and shared energy storage combined system. Where: N is the number of wind turbines in the wind power cluster; P WNi is the rated power of the i-th wind turbine; Δf(t) is the actual frequency of the load;

[0022] f N =50Hz;P e (t) is the actual charging and discharging power of the energy storage;

[0023] The objective function is to minimize the total cost of the wind-storage combined system participating in primary frequency regulation, specifically:

[0024]

[0025] Where: C w1 is the power generation income per unit of electricity; C w2 is the environmental cost per unit of electricity; T f is the frequency modulation time; T is the number of sampling points in one day according to the sampling frequency; C c (r,n) is the capital recovery coefficient; r is the annual interest rate; n is the service life of the energy storage; C e1 is the annual energy storage unit power cost; C e2 is the annual unit capacity cost of energy storage; P E is the rated power of energy storage; E is the capacity of energy storage device; the operation and maintenance cost of energy storage is generally estimated based on a certain proportion of the initial investment, μ is the operation and maintenance cost coefficient of energy storage; C f P is the penalty cost of frequency regulation per unit power;Wr (t) is the frequency regulation capacity that the thermal power primary frequency regulation unit with the same capacity as the wind power cluster should provide at time t;

[0026] The constraints include opportunity constraints, energy storage device charge and discharge power constraints, state of charge continuity constraints, and state of charge constraints. Specifically:

[0027]

[0028] -P E ≤P e (t)≤P E

[0029] SOC t =SOC t-1 +P e (t)×T f

[0030] E×0.1≤SOC≤E×0.9

[0031] Where: P r (·) represents the probability that the equivalent adjustment coefficient of the wind-storage combined system is less than 0.06; β is the given confidence level; P E is the rated charge and discharge power of energy storage; E is the rated capacity of energy storage; SOC is the state of charge of energy storage;

[0032] The above objective functions and constraints together constitute the proposed primary frequency regulation capacity planning model taking into account the joint participation of wind power clusters and shared energy storage.

[0033] Furthermore, in step 4, the chance constraint is simplified according to the pre-given confidence level as follows:

[0034] Where: and Denote the cumulative distribution function of the joint variable and its inverse function, respectively. All constraints in the model are now linear. Using a given confidence level, the Fast Fourier Transform (FFT) is used to quickly calculate the convolution and obtain the probability distribution of the joint variable, which is then quickly solved.

[0035] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0036] 1. Based on the primary frequency regulation of the wind and energy storage system, and taking into account the increasing scale of wind power and the corresponding characteristics of wind power clusters composed of multiple wind turbines, this invention establishes a capacity planning model for the primary frequency regulation of the wind power cluster and shared energy storage system. Compared with traditional wind turbine primary frequency regulation planning, this model realizes the complementarity of wind farm resources, improves the efficiency of energy storage utilization, thereby reducing the required energy storage capacity and achieving better reliability and economy of the wind and energy storage system.

[0037] 2. The present invention provides a solution method for the above-mentioned model, which uses FFT to quickly calculate convolution to obtain the probability distribution of joint variables according to a given confidence level when the number of wind turbines increases and the wind power cluster data is huge, thereby converting chance constraints into deterministic constraints. This avoids the problem of falling into local optimality in the intelligent algorithm based on random simulation, and further improves the efficiency of model solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of the present invention.

[0039] Figure 2 Schematic diagram of the Jensen wake model. DETAILED DESCRIPTION

[0040] A method for primary frequency regulation planning taking into account a wind power cluster and a shared energy storage participating system includes the following steps:

[0041] Step 1, consider the wind speed data that describes the wind speed uncertainty by fitting the probability distribution of the two-parameter Weibull distribution;

[0042] The output characteristic modeling of a wind power cluster simulates the power output characteristics of wind power. First, the probability distribution of the simulated natural wind speed is obtained. Then, the actual wind speed in the wind power cluster is obtained considering the influence of the wind power cluster effect. Finally, the output power of the wind power cluster is obtained by using the wind speed-wind power relationship formula.

[0043] There are many linear shapes commonly used to fit wind speed distribution. In this paper, a two-parameter Weibull distribution is selected to fit wind speed. The distribution function is:

[0044]

[0045] Its probability density function can be expressed as:

[0046]

[0047] Its mean for:

[0048]

[0049] Its variance σ(v 0,t )for:

[0050]

[0051] Where: v 0,t To fit the generated natural wind speed, k and c are the two parameters of the Weibull distribution: k is the shape parameter, and c is the scale parameter. When c = 1, it is called a standard Weibull distribution. Changing the shape parameter k significantly affects the shape of the distribution curve. When k = 1, the distribution is exponential; when k = 2, it is called a Rayleigh distribution; and when k = 3.5, the Weibull distribution is actually very close to a normal distribution. A larger k indicates a wider range of wind speed variation, while a smaller k indicates a smaller wind speed variation. Γ(·) is the gamma function.

[0052] Step 2: Consider the wind power cluster effect including wake effect, delay effect, and topography to establish a wind power cluster output uncertainty model;

[0053] The wind turbine cluster effect includes the wake effect of wind turbines, the wind delay effect, and the influence of the topography of the wind farm. Due to the different locations of wind turbines, the wind speed of wind turbines located upwind will be higher than that of wind turbines located downwind. The closer the distance, the greater the impact. This phenomenon is known as the wake response. The wind delay effect refers to the time it takes for wind to travel from upwind turbines to downwind turbines, i.e., there is a delay.

[0054] The simulation of the wake effect in wind power cluster effect is based on the Jensen wake model, as shown in the attached figure. Figure 2 As shown, when the incoming wind speed is v 0,t , the wind speed at the hub height at a distance l downstream of the wind turbine is v t , the radius of the wind turbine impeller is r0, the wake radius at the downstream position l of the wind turbine is r=r0+sl, s is the velocity recovery coefficient, then the wind speed v of the downstream wind turbine affected by the upstream single wind turbine is t for:

[0055]

[0056] Where: C T is the thrust coefficient, which can generally be taken as 0.2, and s is the speed recovery coefficient. It can be larger for land wind farms, such as 0.075. For offshore wind farms and flat terrain wind farms, the speed recovery is slow, and s is taken as 0.05 or 0.04.

[0057] There are many wind turbines in a wind farm cluster. Downstream wind turbines are often located in the superposition area of the wakes of multiple upstream wind turbines. When downstream wind turbine i is affected by n upstream wind turbines, the kinetic energy loss of the combined wakes of multiple wind turbines is equal to the sum of the kinetic energy losses of the wakes of a single wind turbine. Therefore, the actual wind speed of wind turbine i is:

[0058]

[0059] Where: v i1,t is the wind speed of wind turbine i after considering the wake effect, l ij is the radial distance between wind turbine i and wind turbine j.

[0060] The influence of the topography of the wind farm in the wind power cluster effect is simulated as follows. For a wind farm with undulating terrain, if A is the area covered by the upstream wind turbine wake at the downstream wind turbine location, B is the rotation plane of the downstream wind turbine rotor, and the overlapping part is C. Then the wake shielding area is used as the correction weight, and the wind speed v of wind turbine i after considering the heavy topography of the wind farm cluster is calculated. i2,t for:

[0061]

[0062] The delay effect in wind power clustering is simulated by the delay time of different wind turbines. Most wind turbines in a wind farm are spaced a certain distance apart. It takes a certain delay time for the wind to blow from the first row of wind turbines to the second row of wind turbines. The delay time tl of the wind between the second row of wind turbines can be expressed as:

[0063]

[0064] v i3,t =v i2,t-tl

[0065] Where L is the distance between two adjacent rows of fans along the wind speed direction, v i3,t The wind speed v for the wind power cluster taking into account the wake effect and the influence of topography i2,t The actual wind speed of wind turbine i after the delay time is applied.

[0066] The final wind speed-wind power relationship formula is as follows:

[0067]

[0068] Where: v in is the cut-in wind speed; .v out . is the cut-out wind speed; v r is the rated wind speed; P WNi is the rated power of the i-th wind turbine; P Wi,t is the output power of the i-th wind turbine under different wind speeds, thereby obtaining the wind power cluster output power considering wind speed uncertainty and wind power cluster effect.

[0069] Step 3: Considering the characteristics of multiple wind farms in a wind power cluster scenario, a capacity optimization model for wind power clusters sharing energy storage and jointly participating in system primary frequency regulation based on chance-constrained programming is established;

[0070] In the proposed wind power cluster shared energy storage joint participation in system primary frequency regulation capacity planning, the wind power cluster adopts wind power variable pitch control load shedding operation to realize the shared energy storage frequency regulation and undertake the system primary frequency regulation task together. The primary frequency regulation reserve capacity that the i-th wind turbine in the wind power cluster can provide is:

[0071] P Wri,t =P Wi,t ·x

[0072] Where: x is the load reduction percentage of the wind power cluster. Therefore, the equivalent regulation coefficient R of the wind power cluster and shared energy storage combined system is W for

[0073]

[0074] Where: N is the number of wind turbines in the wind power cluster; P WNi is the rated power of the i-th wind turbine; Δf(t) is the actual frequency of the load; f N =50Hz;P e (t) is the actual charging and discharging power of energy storage.

[0075] The objective function is to minimize the total cost of the wind-storage combined system participating in primary frequency regulation, specifically:

[0076]

[0077] Where: C w1 is the power generation income per unit of electricity; C w2 is the environmental cost per unit of electricity; T f is the frequency modulation time; T is the number of sampling points in one day according to the sampling frequency; C c (r,n) is the capital recovery coefficient; r is the annual interest rate, which is 0.1; n is the energy storage service life, which is 10 years; C e1 is the annual energy storage unit power cost; C e2 is the annual unit capacity cost of energy storage; P E is the rated power of the energy storage; E is the rated capacity of the energy storage device; the operation and maintenance costs of the energy storage are generally estimated based on a certain proportion of the initial investment, μ is the operation and maintenance cost coefficient of the energy storage, which is taken as 0.5%; C f P is the penalty cost of frequency regulation per unit power; Wr (t) is the frequency regulation capacity that the thermal power primary frequency regulation unit with the same capacity as the wind power cluster should provide at time t.

[0078] Constraints include opportunity constraints, energy storage device charge and discharge power constraints, state of charge continuity constraints, and state of charge constraints. Specifically:

[0079]

[0080] -P E ≤P e (t)≤P E

[0081] SOC t =SOC t-1 +P e (t)×T f

[0082] E×0.1≤SOC≤E×0.9

[0083] Where: P r (·) represents the probability that the equivalent adjustment coefficient of the wind-storage combined system is less than 0.06; β is the given confidence level; P E is the rated power of energy storage; E is the rated capacity of energy storage device; SOC is the state of charge of energy storage.

[0084] The above objective functions and constraints together constitute the proposed primary frequency regulation capacity planning model taking into account the joint participation of wind power clusters and shared energy storage.

[0085] Step 4: With a given confidence level, the convolution is quickly calculated using the Fast Fourier Transform (FFT) to obtain the probability distribution of the joint variables, formally transforming the chance constraint into a deterministic constraint, and then solving it to obtain the optimal capacity configuration plan for the wind power cluster load reduction rate and shared energy storage.

[0086] The chance constraint is simplified according to a given confidence level:

[0087]

[0088]

[0089] Where: and Represent the cumulative distribution function of the random variable and its inverse function respectively. So far, all constraints in the model are linear constraints. When the sequence of discrete convolution is very long, it will also consume a lot of time. According to the convolution in the time domain is equivalent to the product in the frequency domain, FFT is introduced to quickly calculate the convolution to obtain the probability distribution of the joint variable. The computational complexity of FFT is O(N·logN), while the complexity of convolution operation is O(N 2 ), thus speeding up the solution process. Using FFT to quickly obtain the probability distribution of joint variables can be summarized as follows: first, the discrete PDF arrays of each random variable are converted to the frequency domain using FFT. The resulting frequency domain arrays are then multiplied continuously, and then inversely transformed to the time domain using IFFT to obtain the convolution result. Finally, the CPLEX solver is called using MATLAB's YALMIP, making the solution more convenient.

Claims

1. A method for primary frequency regulation capacity planning taking into account the joint participation of wind power cluster shared energy storage, characterized by: The following steps are involved: Step 1, fitting and generating wind speed data describing wind speed uncertainty; Step 2: Considering the wind turbine cluster effect, including the wake effect, delay effect, and topographic influence, a wind turbine cluster output uncertainty model is established. Step 2 specifically includes: the wind turbine cluster effect includes the wind turbine wake effect, wind delay effect, and the influence of the topography where the wind farm is located. The simulation of the wake effect in wind power cluster effect is based on the Jensen wake model. When the incoming wind speed is v 0,t , the wind speed at the hub height at a distance l downstream of the wind turbine is v t , the radius of the wind turbine impeller is r0, the wake radius at the downstream position l of the wind turbine is r=r0+sl, s is the velocity recovery coefficient, then the wind speed v of the downstream wind turbine affected by the upstream single wind turbine is t for: Where: C T is the thrust coefficient, s is the speed recovery coefficient; There are many wind turbines in a wind farm cluster. The downstream wind turbines are located in the superposition area of the wakes of multiple upstream wind turbines. When downstream wind turbine i is affected by n upstream wind turbines, the kinetic energy loss of the mixed wakes of multiple wind turbines is equal to the sum of the kinetic energy losses of the wakes of a single wind turbine. Therefore, the actual wind speed of wind turbine i is: Where: v i1,t is the wind speed of wind turbine i after considering the wake effect, l ij is the radial distance between wind turbine i and wind turbine j; The influence of the topography of the wind farm in the wind power cluster effect is simulated as follows. For a wind farm with undulating terrain, if A is the coverage area of the upstream wind turbine wake at the downstream wind turbine location, B is the rotation plane of the downstream wind turbine rotor, and the overlapping part is C, then the wake shielding area is used as the correction weight. After considering the influence of the wind power cluster's topography, the wind speed v of wind turbine i is i2,t for: The delay effect in the wind power cluster effect is simulated by the delay time of different wind turbines. The wind turbines in the wind farm are a certain distance apart. It takes a certain delay time for the wind to blow from the first row of wind turbines to the second row of wind turbines. The delay time tl of the wind between the second row of wind turbines can be expressed as: v i3,t =v i2,t-tl , where L is the distance between two adjacent rows of fans along the wind speed direction, v i3,t The wind speed v for the wind power cluster taking into account the wake effect and the influence of topography i2,t The actual wind speed of wind turbine i after the delay time is applied; The final wind speed-wind power relationship formula is as follows: Where: v in is the cut-in wind speed; v out is the cut-out wind speed; v r is the rated wind speed; P WNi is the rated capacity of the i-th wind turbine; P Wi,t is the output power of the i-th wind turbine under different wind speeds, thereby obtaining the wind power cluster output power considering wind speed uncertainty and wind power cluster effect; Step 3: Considering the characteristics of multiple wind farms in a wind power cluster scenario, a capacity planning model for wind power clusters sharing energy storage and jointly participating in system primary frequency regulation based on chance-constrained programming is established. Step 4: With a given confidence level, the convolution is quickly calculated using the Fast Fourier Transform (FFT) to obtain the probability distribution of the joint variables, formally transforming the chance constraint into a deterministic constraint, and then solving it to obtain the optimal capacity configuration plan for the wind power cluster load reduction rate and shared energy storage.

2. The method for primary frequency regulation capacity planning taking into account the joint participation of wind power cluster shared energy storage according to claim 1, characterized in that: Step 1 is as follows: Select the two-parameter Weibull distribution to fit the wind speed, and the distribution function is: Its probability density function can be expressed as: Its mean for: Its variance σ(v 0,t )for: Where: v 0,t To fit the generated natural wind speed, k and c are two parameters of the Weibull distribution, k is the shape parameter, and c is the scale parameter.

3. The method for primary frequency regulation capacity planning taking into account the joint participation of wind power cluster shared energy storage according to claim 2, characterized in that: Step 3 is as follows: In the planning of the primary frequency regulation capacity of the wind power cluster and the shared energy storage, the wind power cluster adopts wind power variable pitch control load shedding operation to realize the shared primary frequency regulation task of the system together with the shared energy storage frequency regulation. The primary frequency regulation reserve capacity that the i-th wind turbine in the wind power cluster can provide is: P Wri,t =P Wi,t x, where x is the load reduction percentage of the wind power cluster and the equivalent regulation coefficient of the wind power cluster and shared energy storage combined system. Where: N is the number of wind turbines in the wind power cluster; P WNi is the rated power of the i-th wind turbine; Δf(t) is the actual frequency of the load; f N =50Hz;P e (t) is the actual charging and discharging power of energy storage; the objective function is to minimize the total cost of the wind-storage combined system participating in primary frequency regulation, specifically: Where: C w1 is the power generation income per unit of electricity; C w2 is the environmental cost per unit of electricity; T f is the frequency modulation time; T is the number of sampling points in one day according to the sampling frequency; C c (r,n) is the capital recovery coefficient; r is the annual interest rate; n is the service life of the energy storage; C e1 is the annual energy storage unit power cost; C e2 is the annual unit capacity cost of energy storage; P E is the rated power of energy storage; E is the capacity of energy storage device; the operation and maintenance cost of energy storage is generally estimated based on a certain proportion of the initial investment, μ is the operation and maintenance cost coefficient of energy storage; C f P is the penalty cost of frequency regulation per unit power; Wr (t) is the frequency regulation capacity that the thermal power primary frequency regulation unit with the same capacity as the wind power cluster should provide at time t; The constraints include opportunity constraints, energy storage device charge and discharge power constraints, state of charge continuity constraints, and state of charge constraints. Specifically: -P E ≤P e (t)≤P E SOCIETY t =SOC t-1 +P e (t)×T f E×0.1≤SOC≤E×0.9 Where: P r (·) represents the probability that the equivalent adjustment coefficient of the wind-storage combined system is less than 0.06; β is the given confidence level; P E is the rated charge and discharge power of energy storage; E is the rated capacity of energy storage; SOC is the state of charge of energy storage; The above objective functions and constraints together constitute the proposed primary frequency regulation capacity planning model taking into account the joint participation of wind power clusters and shared energy storage.

4. The method for primary frequency regulation capacity planning taking into account the joint participation of wind power cluster shared energy storage according to claim 3, characterized in that: In step 4, the chance constraint is simplified according to the pre-given confidence level as follows: Where: and They represent the cumulative distribution function of the joint variable and its inverse function respectively. So far, all constraints in the model are linear constraints. With a given confidence level, the fast Fourier transform (FFT) is used to quickly calculate the convolution to obtain the probability distribution of the joint variable and then quickly solve it.

Images