A wind farm fast frequency response method based on a distributed control strategy

Abstract

A kind of wind farm fast frequency response method based on distributed control strategy.The present application relates to the technical field of wind farm frequency optimization, in particular to the technical field of distributed control wind farm frequency response optimization.The present application effectively solves the problems of slow wind farm frequency response speed, poor robustness and high control complexity, and improves the efficiency of wind farm frequency response optimization.The method comprises the following steps: establishing a wind farm partition model, dividing multiple wind turbines into independent control units;Determine the optimal classification threshold and confirm the optimal fuzzy classification;Problem modeling;Problem decomposition;Iterative optimization.The present application has a significant difference from the prior art in improving the speed and flexibility of wind farm frequency response, and solves the problems of response delay, high control complexity and insufficient system stability in the prior art.

Description

Technical Field

[0001] This invention relates to the field of wind farm frequency optimization technology, and more specifically to the field of distributed control wind farm frequency response optimization technology. Background Technology

[0002] With the continuous growth of global demand for renewable energy, wind power, as an important component of renewable energy, is seeing its application scope and scale expand. However, the randomness and intermittency of wind power pose a significant challenge to the frequency stability of power systems. Traditional power systems mainly rely on conventional power generation methods such as thermal power and hydropower to regulate frequency, but these methods often fall short when faced with large-scale wind power integration, and the speed and stability of frequency response are difficult to meet requirements.

[0003] When wind farms are connected to the grid, grid frequency fluctuations intensify. Traditional centralized control methods for frequency regulation typically rely on a single central controller for global control. This approach faces the following main challenges:

[0004] Slow response speed: Due to communication delays and complex control structures, centralized control is not fast enough to respond to frequency changes in large-scale wind farms.

[0005] Poor system robustness: As the core of the system, if the central controller fails, the frequency response of the entire system will be severely affected.

[0006] High control complexity: As the scale of wind farms expands, the implementation of centralized control strategies becomes more complex and unstable. Summary of the Invention

[0007] To address the aforementioned problems, this invention discloses a fast frequency response method for wind farms based on a distributed control strategy, relating to the field of wind farm frequency optimization technology. It effectively solves the problems of slow frequency response speed, poor robustness, and high control complexity in wind farms, thereby improving the efficiency of wind farm frequency response optimization.

[0008] The method includes the following steps:

[0009] S1. Establish a wind farm zoning model and divide multiple wind turbines into independent control units;

[0010] S2. Determine the optimal classification threshold and confirm the optimal fuzzy classification;

[0011] S3, Problem Modeling;

[0012] S4. Problem Breakdown;

[0013] S5. Iterative optimization.

[0014] Furthermore, the establishment of the wind farm zoning model specifically involves: using fuzzy clustering to zonify the wind farm, dividing multiple wind turbines into independent control units; the fuzzy clustering analysis method specifically involves: constructing a feature matrix based on the electrical distance between nodes, constructing a fuzzy similarity matrix based on the feature matrix, and then performing fuzzy clustering based on the fuzzy similarity matrix.

[0015] Furthermore, the electrical distance between the nodes is calculated using their physical locations and electrical parameters; the feature matrix is ​​specifically: X = (x ij ) n×m The x ij Let x represent the electrical distance between node i and node j, where x represents the electrical distance, n represents the number of features, and m represents the total number of nodes.

[0016] Furthermore, the fuzzy similarity matrix is ​​specifically as follows: Where c represents the optimal number of clusters, k represents the index variable, k = [1, 2, 3, ..., m], and x ik The x represents the value of node i at the k-th feature value. jk This represents the value of node j at the kth feature value; the fuzzy clustering specifically involves obtaining the fuzzy equivalence matrix from the fuzzy similarity matrix using the successive squaring method.

[0017] Furthermore, the optimal classification threshold is determined by... Confirmed; the c represents the average value of the i-th type node on the k-th feature value. i Representing the i-th class, the As shown This represents the number of nodes in the i-th class, where j represents the j-th node in the i-th class; when and When they are equal, the current fuzzy classification is the best fuzzy classification. This represents the average value of the k-th feature across all nodes.

[0018] Furthermore, the problem modeling specifically involves: modeling the global frequency response problem as a global optimization problem, wherein the global optimization problem specifically includes: The f i () represents the local objective function of the i-th wind turbine, where w i Let represent the frequency response parameter of the i-th wind turbine, let represent the linear constraint condition, let A represent an m×n dimensional matrix, let w represent an n×1 column vector, and let b represent the coupling constraint of the system.

[0019] Furthermore, the problem decomposition specifically involves: decomposing the global optimization problem into subproblems using the alternating direction multiplier method; the subproblems are specifically: The λ T Let λ denote the transpose of the Lagrange multiplier, ρ denote the coefficient of the penalty term, and ||()|| 2 This represents the square of the norm.

[0020] Furthermore, the iterative optimization specifically involves: optimizing each wind turbine independently of its sub-problem and updating w. i According to w i Update the global variable z, and then update the Lagrange multiplier λ based on the global variable z; the update w i Specifically: The This indicates that w represents the value at the (k+1)th iteration. i The updated value, the f i (w i When minimizing w i The value of z k This represents the global variable for the k-th iteration; the statement based on w i Updating the global variable z specifically involves: The z k+1 This represents the updated value of z in the (k+1)th iteration; the update of the Lagrange multipliers based on the global variable z specifically refers to: λ k+1 =λ k +ρ(w k+1 -z k+1 ), the λ k+1 Let λ represent the Lagrange multiplier in the (k+1)th iteration. k Denotes the Lagrange multiplier in the k-th iteration; when When the optimization iteration ends, the ||Aw k -b|| 2 Represents the original residual, w k Describes the variable vector for the k-th iteration, the This indicates the preset precision threshold.

[0021] The beneficial effects of this invention are as follows:

[0022] This invention proposes a fast frequency response method for wind farms based on a distributed control strategy. This method is innovative in its control strategy and specifically optimizes the wind farm's rapid response to grid frequency fluctuations. The distributed control strategy allows wind turbines within the wind farm to respond autonomously based on local conditions. Simultaneously, through information exchange and collaborative control between modules, it ensures consistent and efficient frequency response across the entire farm. This method improves response speed while maintaining system stability control, avoiding single points of failure and communication delays that may exist in existing centralized control systems. This invention significantly differs from existing technologies in improving the frequency response speed and flexibility of wind farms, solving problems such as response hysteresis, high control complexity, and insufficient system stability in existing methods. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the zoning process of a wind farm according to an embodiment of the present invention;

[0024] Figure 2 This is a flowchart illustrating the iterative optimization process of an embodiment of the present invention;

[0025] Figure 3 This is a schematic diagram of dynamic clustering in an embodiment of the present invention;

[0026] Figure 4 This is a schematic diagram illustrating the coordination and optimization of an independent control unit in an embodiment of the present invention. Detailed Implementation

[0027] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] This embodiment provides a fast frequency response method for wind farms based on a distributed control strategy, the method comprising the following steps:

[0029] S1. Establish a wind farm zoning model and divide multiple wind turbines into independent control units.

[0030] The relevant operations in step S1 will be introduced with specific examples:

[0031] In large-scale wind farms, distributed control strategies can be used to coordinate the frequency responses of individual wind turbines to achieve overall system optimization. For example... Figure 1As shown, firstly, the wind farm is divided into multiple independent control units using a clustering algorithm, and each unit can be optimized for frequency response independently. Considering that hard partitioning methods may not be suitable for objects with fuzzy or changing features, fuzzy clustering analysis can more reasonably divide the control area, thereby improving the system's response efficiency and accuracy.

[0032] The fuzzy clustering analysis method specifically involves: constructing a feature matrix based on the electrical distance between nodes, constructing a fuzzy similarity matrix based on the feature matrix, and then performing fuzzy clustering based on the fuzzy similarity matrix.

[0033] Each wind farm has n features, and each wind farm can be represented by a vector containing m nodes, a process called calibration. Calibration involves quantifying each feature of the wind farm to be classified into specific data. The electrical distance between nodes is calculated using their physical location and electrical parameters. The feature matrix is ​​specifically: X = (x... ij ) n×m The x ij x represents the electrical distance between node i and node h.

[0034] The nodes i and j represent node indices, characterized by the electrical distance between each node and other nodes. The feature matrix of the wind farm to be classified is composed of the electrical distances between nodes. Optimizing the classification process requires fully considering these electrical distance features to ensure accurate identification and classification when analyzing the relationships between nodes, thereby improving the overall system performance and efficiency. To measure the similarity between the wind farms to be classified, an absolute value subtraction method is used to construct a fuzzy similarity matrix to analyze the similarity between nodes.

[0035] The fuzzy similarity matrix is ​​specifically as follows: The c represents the optimal clustering number, such that r ij The values ​​of x are scattered between [0, 1], where k represents the index variable, k = [1, 2, 3, ..., m], and x ik The x represents the value of node i at the k-th feature value. jk The value of node j at the kth eigenvalue is represented; the fuzzy clustering is specifically performed by: obtaining the transitive closure matrix of the fuzzy similarity matrix through the successive squaring method, wherein the transitive closure matrix is ​​a fuzzy equivalence matrix.

[0036] Fuzzy clustering analysis is a classification method based on fuzzy mathematics. It classifies objective things by analyzing their different characteristics, degree of closeness, and similarity, and can effectively handle classification problems with fuzziness and uncertainty.

[0037] S2. Determine the optimal classification threshold and confirm the optimal fuzzy classification.

[0038] The relevant operations in step S2 will be introduced with specific examples:

[0039] The optimal classification threshold is obtained through Confirmed; the c represents the average value of the i-th type node on the k-th feature value. i Representing the i-th class, the As shown This represents the number of nodes in the i-th class, where j represents the j-th node in the i-th class; for example... Figure 3 As shown, the fuzzy equivalence matrix is ​​truncated according to an appropriate threshold. This appropriate threshold is adjusted dynamically by observing fuzzy clustering results, optimizing intra-cluster compactness and inter-cluster separation, calculating classification error, and considering application scenarios and experimental results. When r ij As the value decreases from 1 to 0, the fuzzy classification gradually merges from fine to coarse, forming a dynamic clustering graph. This dynamic clustering graph represents the dynamic clustering result. and When they are equal, the current fuzzy classification is the best fuzzy classification. The value represents the average of the k-th feature among all nodes.

[0040] S3, Problem Modeling.

[0041] The relevant operations in step S3 will be introduced with specific examples:

[0042] Wind farms involve various complex constraints, including: power output limits for wind turbines, frequency security constraints of the power grid, and limitations on the response speed and capacity of wind turbines. For example... Figure 2 As shown, the Distributed Optimization Model (ADMM) decomposes constraints, mapping global constraints to each local subproblem. This allows wind turbines to optimize independently while satisfying their own constraints, and enables them to flexibly handle various complex constraints in the system. With the objective of minimizing the frequency deviation of the entire power system by adjusting the frequency response of each wind turbine, while simultaneously satisfying the constraints of both the wind farm and the wind turbines, the global frequency response problem is modeled as a global optimization problem. Specifically, the global optimization problem is as follows: The f i () represents the local objective function of the i-th wind turbine, where w i Let represent the frequency response parameter of the i-th wind turbine, let represent the linear constraint condition, let A represent an m×n dimensional matrix, let w represent an n×1 column vector, and let b represent the coupling constraint of the system.

[0043] This distributed architecture significantly improves computational scalability, making it suitable for large-scale wind farms and systems composed of multiple wind farms. The algorithm remains highly efficient as the number of wind farms increases, without a significant performance degradation due to system size growth. A communication network supports distributed optimization, with the distributed optimization algorithm coordinating and optimizing through data exchange between the scheduling center and each wind turbine.

[0044] S4. Problem Breakdown.

[0045] The relevant operations in step S4 will be introduced with specific examples:

[0046] Distributed optimization algorithms can decompose a global optimization problem into subproblems for multiple wind turbines, and each wind turbine can process the subproblems in parallel.

[0047] The problem decomposition specifically involves: decomposing the global optimization problem into subproblems using the alternating direction multiplier method; the subproblems are specifically: The λ T Let λ denote the transpose of the Lagrange multiplier, ρ denote the coefficient of the penalty term, and ∥()∥ 2 This represents the square of the norm.

[0048] S5. Iterative optimization.

[0049] The relevant operations in step S5 will be introduced with specific examples:

[0050] The iterative optimization specifically involves: optimizing the sub-problem independently for each wind turbine and updating w. i According to w i Update the global variable z, and then update the Lagrange multiplier λ based on the global variable z; the update w i Specifically: The This indicates that w represents the value at the (k+1)th iteration. i The updated value, the f i (w i When minimizing w i The value of z k This represents the global variable for the k-th iteration; the statement based on w i Updating the global variable z specifically involves: The z k+1 This represents the updated value of z in the (k+1)th iteration; the update of the Lagrange multipliers based on the global variable z specifically refers to: λ k+1 =λ k +ρ(w k+1 -z k+1 ), the λ k+1 Let λ represent the Lagrange multiplier in the (k+1)th iteration. kDenotes the Lagrange multiplier in the k-th iteration; when When the optimization iteration ends, the ||Aw k -b|| 2 Represents the original residual, w k Describes the variable vector for the k-th iteration, the This indicates the preset precision threshold.

[0051] like Figure 4 As shown, each wind turbine can independently adjust its output power based on the local grid frequency and real-time operating status. This design enables the turbines to respond quickly to frequency fluctuations, and because each turbine is a relatively independent control unit, it no longer relies on a single central controller, thus significantly reducing communication latency issues and the bottleneck of centralized control. This autonomous response design not only improves response speed but also enhances the system's fault tolerance; even if some turbines fail, the others can continue to operate, maintaining the stability of the overall frequency response.

[0052] Although each wind turbine can adjust autonomously, ensuring a consistent frequency response across the entire wind farm requires information exchange and coordinated control among the turbines. This necessitates a communication network between the turbines, allowing them to share crucial information such as real-time grid frequency and local operating status. This information sharing enables not only autonomous optimization but also collaborative optimization based on the overall grid status, ensuring consistent overall frequency response.

Claims

1. A fast frequency response method for wind farms based on a distributed control strategy, characterized in that, The method includes the following steps: S1. Establish a wind farm zoning model and divide multiple wind turbines into independent control units; S2. Determine the optimal classification threshold and confirm the optimal fuzzy classification; S3, Problem Modeling; The problem modeling specifically involves: modeling the global frequency response problem as a global optimization problem, wherein the global optimization problem is specifically: The Indicates the first The local objective function of each wind turbine, the Indicates the first The frequency response parameters of the wind turbine, the Represents the linear constraint condition, the Represent a dimensional matrix, the Represent a The column vector, the This represents the coupling constraints of the system; S4. Problem Breakdown; The problem decomposition specifically involves: decomposing the global optimization problem into subproblems using the alternating direction multiplier method; the subproblems are specifically: The Denotes the transpose of the Lagrange multiplier, the Describing the Lagrange multipliers, the stated The coefficient representing the penalty, the Represents the square of the norm; S5, Iterative optimization; The iterative optimization specifically involves: optimizing each wind turbine's sub-problem independently and updating... ,according to Update global variables Then based on global variables Update Lagrange multipliers The update Specifically: The Indicates the first During the next iteration The updated value, the express When minimized The value, the Indicates the first The global variable for the next iteration; the basis Update global variables Specifically: The Indicates the first During the next iteration The updated value; based on global variables The specific steps for updating the Lagrange multipliers are as follows: The Indicates the first The Lagrange multipliers of the next iteration, the Indicates the first The Lagrange multipliers of the next iteration; when When the optimization iteration ends, the aforementioned Represents the original residual, the stated Indicates the first The variable vector of the next iteration, the This indicates the preset precision threshold.

2. The fast frequency response method for wind farms based on a distributed control strategy according to claim 1, characterized in that, The establishment of the wind farm zoning model specifically involves: using fuzzy clustering analysis to zonify the wind farm, dividing multiple wind turbines into independent control units; the fuzzy clustering analysis method specifically involves: constructing a feature matrix based on the electrical distance between nodes, constructing a fuzzy similarity matrix based on the feature matrix, and then performing fuzzy clustering based on the fuzzy similarity matrix.

3. The fast frequency response method for wind farms based on a distributed control strategy according to claim 2, characterized in that, The electrical distance between the nodes is calculated using their physical locations and electrical parameters; the feature matrix is ​​specifically: The Represents a node and nodes The electrical distance between them, the Indicates electrical distance, the Indicates the number of features, the This indicates the total number of nodes.

4. The fast frequency response method for wind farms based on a distributed control strategy according to claim 3, characterized in that, The fuzzy similarity matrix is ​​specifically as follows: The Denotes the optimal number of clusters, the Represents an index variable, the The Represents a node In the The values ​​taken at each eigenvalue, the Represents a node In the The value of each feature; the fuzzy clustering is specifically performed by taking the fuzzy similarity matrix as a fuzzy equivalence matrix obtained by the successive squaring method.

5. The fast frequency response method for wind farms based on a distributed control strategy according to claim 4, characterized in that, The optimal classification threshold is obtained through Confirmed; the Indicates the first Class nodes in the The average of the values ​​taken over the eigenvalues, the Indicates the first The class, the As shown Indicates the first Number of class nodes, the Indicates the first The class of 1 node; when and When they are equal, the current fuzzy classification is the best fuzzy classification. This represents the average value of the k-th feature across all nodes.

Images