A wind turbine and a wind farm intelligent optimization control system and method

Abstract

The application discloses a wind turbine and a wind farm intelligent optimization control system and method, relates to the technical field of wind farm optimization control, and combines abnormality degree with a clustering algorithm to complete clustering processing of residual wind turbines, carries out clustering processing on the residual wind turbines through the abnormality degree, corrects the maximum limit rotating speed of the wind turbines in each category, monitors the real-time rotating speed of the wind turbines in each category, dynamically adjusts the pitch angle of the wind turbine blades based on the comparison result of the real-time rotating speed and the maximum limit rotating speed, generates a wind farm electric energy control strategy after analyzing the real-time power generation of the wind farm and the real-time demand power of the power grid, and the control system can effectively improve the coordination between the wind turbines, makes the wind turbine group operate more efficiently, avoids the negative influence of the poor operation of individual wind turbines on the overall power generation efficiency, generates the wind farm electric energy control strategy through the analysis of the real-time power generation of the wind farm and the real-time demand power of the power grid, and improves the cooperative working efficiency of the wind farm and the power grid.

Description

Technical Field

[0001] This invention relates to the field of wind farm optimization control technology, specifically to a wind turbine and a wind farm intelligent optimization control system and method. Background Technology

[0002] A wind farm is a facility that generates electricity using wind energy. It typically consists of multiple wind turbines that capture the kinetic energy of the wind and convert it into electrical energy, providing clean energy for the power system. With the increasing global demand for renewable energy, wind farms have become an important part of the development of green energy.

[0003] The existing technology has the following shortcomings:

[0004] 1. Wind turbines in a wind farm usually operate independently, and the coordination and optimal configuration between turbines are not fully considered. When wind speed and direction change, some turbines may operate inefficiently due to their location or uneven wind speed, resulting in a decrease in the power generation efficiency of the entire wind farm.

[0005] 2. Traditional wind farm control systems often only focus on the operating status of a single wind turbine, lacking coordination and optimization between the power generation of the entire wind farm and the grid demand. This can easily lead to the wind farm outputting too much or too little power, failing to accurately match the grid demand, resulting in resource waste or grid instability.

[0006] This invention provides a wind turbine and a wind farm intelligent optimization control system and method, which can effectively improve the coordination between wind turbines, make the wind turbine group operate more efficiently, avoid the negative impact of individual wind turbines operating poorly on the overall power generation efficiency, and generate a wind farm power control strategy by analyzing the real-time power generation of the wind farm and the real-time power demand of the grid, thereby improving the collaborative working efficiency between the wind farm and the grid. Summary of the Invention

[0007] The purpose of this invention is to provide a wind turbine and a wind farm intelligent optimization control system and method to address the shortcomings in the prior art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a wind turbine for a wind farm and a smart optimization control method for a wind farm, the optimization method comprising the following steps:

[0009] S1: The control system obtains wind farm information based on the API interface of the wind farm management platform, and numbers all wind turbines in the wind farm according to their location. During the operation of the wind farm, the system monitors the operating data of each wind turbine in real time and calculates the anomaly degree of the wind turbine through regression analysis algorithm.

[0010] S2: First, shut down the wind turbines that do not support operation based on the anomaly. Then, combine the anomaly with the clustering algorithm to complete the clustering of the remaining wind turbines. Correct the maximum speed limit of the wind turbines in each category and monitor the real-time speed of the wind turbines in each category. Based on the comparison between the real-time speed and the maximum speed limit, dynamically adjust the blade pitch angle of the wind turbines.

[0011] S3: After analyzing the real-time power generation of the wind farm and the real-time power demand of the power grid, generate a wind farm power control strategy.

[0012] In a preferred embodiment, after real-time monitoring of the operating data of each wind turbine, the anomaly degree of the wind turbine is calculated using a regression analysis algorithm, including the following steps:

[0013] Monitor the wind turbine's operating data, including blade curvature index, generator temperature and noise rise rate, gearbox vibration frequency, and main shaft strain factor;

[0014] The abnormality of the wind turbine is calculated by substituting the blade bending index, generator temperature and noise rise rate, gearbox vibration frequency, and main shaft strain factor into the regression analysis algorithm.

[0015] In a preferred embodiment, the clustering of remaining wind turbines is completed by combining anomaly detection with a clustering algorithm, including the following steps:

[0016] S2.1: Obtain the remaining support turbine numbers and quantities, and generate an initial K value (i.e., the number of clusters) based on the remaining turbine numbers. In the formula, N represents the number of remaining wind turbines. Indicates rounding down;

[0017] S2.2: Randomly calculate the anomaly of K wind turbines as the initial cluster center value, and calculate the difference between the remaining wind turbines and each initial cluster center value. Assign the remaining wind turbines to the cluster with the smallest difference.

[0018] S2.3: After all remaining wind turbines have been allocated, calculate the mean anomaly of each cluster as the new cluster center value;

[0019] S2.4: Repeat steps S2.2 and S2.3. When the maximum number of iterations is reached or the change in the cluster center value is less than the change threshold, it is determined that the convergence condition is met, and K clusters are output.

[0020] In a preferred embodiment, the maximum speed limit of the fan in each category is corrected, including the following steps:

[0021] After obtaining the K clusters and their cluster center values, the maximum speed limit of the fan in each cluster is adjusted based on the cluster center values. The expression is as follows:

[0022] rsp correct =rsp initial *(1-unusual avg In the formula, rsp correct To correct the maximum speed limit, rsp initial The initial maximum speed limit, unusual avg Let be the cluster center value of the cluster, and In the formula, N represents the number of remaining wind turbines, unusual i Let be the anomaly degree of the i-th remaining wind turbine.

[0023] In a preferred embodiment, correcting the maximum limiting speed of the wind turbine in each category further includes the following steps:

[0024] The anomaly degree of supporting the operation of the remaining wind turbines is compared with the first anomaly threshold. The first anomaly threshold is less than the second anomaly threshold. The first anomaly threshold is used to analyze whether the maximum speed limit of the wind turbines needs to be corrected.

[0025] If the abnormality of the fan is less than or equal to the first abnormality threshold, the analysis does not require correction of the maximum speed limit of the fan.

[0026] If the anomaly of a fan exceeds the first anomaly threshold, the analysis requires correction to the maximum speed limit of that fan. This correction is based on the cluster center value, adjusting the maximum speed limit of the fans in each cluster using the following expression: rsp correct =rsp initial *(1-unusual avg In the formula, rsp correct To correct the maximum speed limit, rsp initial The initial maximum speed limit, unusual avg Let be the cluster center value of the cluster, and In the formula, N represents the number of remaining wind turbines, unusual i Let be the anomaly degree of the i-th remaining wind turbine.

[0027] In a preferred embodiment, dynamically adjusting the blade pitch angle of the wind turbine based on a comparison between the real-time rotational speed and the maximum limiting rotational speed includes the following steps:

[0028] During the operation of the wind farm, the real-time speed of each internal wind turbine is monitored. If the real-time speed of the wind turbine is less than the corrected maximum limit speed, the pitch angle of the wind turbine blades is increased until the real-time speed of the wind turbine is equal to the corrected maximum limit speed.

[0029] If the real-time speed of the wind turbine is greater than the corrected maximum speed limit, the blade pitch angle of the wind turbine is reduced until the real-time speed of the wind turbine equals the corrected maximum speed limit.

[0030] In a preferred embodiment, shutting down unsupported fans based on the degree of abnormality includes the following steps:

[0031] The higher the anomaly level of the fan, the more severe the anomaly, and the less support the fan is for operation. The obtained anomaly level is compared with a preset second anomaly threshold. The second anomaly threshold is used to analyze whether the fan can support operation. If the fan's anomaly level is less than or equal to the second anomaly threshold, the fan is analyzed to be normal and can support operation. If the fan's anomaly level is greater than the second anomaly threshold, the fan is analyzed to be abnormal, cannot support operation, and needs to be shut down.

[0032] In a preferred embodiment, the calculation logic for the blade curvature index is as follows: obtain the number of blades on the wind turbine and calculate the curvature of each blade, expressed as: In the formula, θ(x) represents the degree of blade curvature at the detection point x, E is the elastic modulus of the blade material, I is the moment of inertia of the blade cross section, and M(x) is the bending moment acting at the detection point x. The blade curvature index of the wind turbine is calculated based on the degree of curvature of each blade, and the expression is: In the formula, Let θ(x) be the blade curvature index, n be the number of blades on the wind turbine, and θ(x) be the blade bending index. i Let x be the degree of curvature of the i-th blade at the detection point x.

[0033] In a preferred embodiment, the expression for calculating the bending moment acting at the detection point x is: In the formula, M(x) is the bending moment acting at the detection point x, and W is the weight of the blade. Let x represent the distance from the point of application of gravity to the detection point, and L be the length of the blade. The expression for calculating the weight of the blade is: In the formula, W is the weight of the blade, ρ mat is the density of the blade material, A(x) is the cross-sectional area of ​​the blade at the detection point x, g is the gravitational acceleration, and L is the length of the blade.

[0034] A wind farm turbine and wind farm intelligent optimization control system includes a numbering module, a clustering module, a dynamic control module, and a strategy generation module.

[0035] Numbering module: Based on the API interface of the wind farm management platform, wind farm information is obtained, and all wind turbines in the wind farm are numbered sequentially according to their location.

[0036] Clustering module: During the operation of the wind farm, after real-time monitoring of the operating data of each wind turbine, the anomaly degree of the wind turbine is calculated by regression analysis algorithm. First, the wind turbines that do not support operation are shut down based on the anomaly degree. Then, the anomaly degree is combined with the clustering algorithm to complete the clustering process of the remaining wind turbines.

[0037] Dynamic control module: Monitors the real-time speed of the wind turbines in each category, and dynamically adjusts the blade pitch angle based on the comparison between the real-time speed and the maximum limit speed, thereby controlling the wind-receiving area of ​​the blades;

[0038] Strategy generation module: After analyzing the real-time power generation of the wind farm and the real-time power demand of the power grid, it generates the power control strategy for the wind farm.

[0039] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0040] This invention calculates the anomaly degree of wind turbines using regression analysis algorithms, then combines the anomaly degree with a clustering algorithm to cluster the remaining wind turbines. After clustering the remaining turbines using the anomaly degree, the maximum speed limit of the turbines in each category is corrected, and the real-time speed of the turbines in each category is monitored. Based on the comparison between the real-time speed and the maximum speed limit, the pitch angle of the turbine blades is dynamically adjusted. After analyzing the real-time power generation of the wind farm and the real-time power demand of the grid, a wind farm power control strategy is generated. The control system effectively improves the coordination between wind turbines, making the operation of the wind turbine group more efficient and avoiding the negative impact of individual turbine malfunctions on overall power generation efficiency. By analyzing the real-time power generation of the wind farm and the real-time power demand of the grid, the wind farm power control strategy is generated, improving the collaborative working efficiency between the wind farm and the grid. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0042] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. 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.

[0044] Example 1: Please refer to Figure 1 As shown in this embodiment, a wind turbine and a wind farm intelligent optimization control method for wind farms are described. The optimization method includes the following steps:

[0045] S1: The control system obtains wind farm information based on the API interface of the wind farm management platform, and numbers all wind turbines in the wind farm according to their location. During the operation of the wind farm, the system monitors the operating data of each wind turbine in real time and calculates the anomaly degree of the wind turbine through regression analysis algorithm.

[0046] S2: First, shut down the wind turbines that do not support operation based on the anomaly. Then, combine the anomaly with the clustering algorithm to complete the clustering of the remaining wind turbines. Correct the maximum speed limit of the wind turbines in each category and monitor the real-time speed of the wind turbines in each category. Based on the comparison between the real-time speed and the maximum speed limit, dynamically adjust the blade pitch angle of the wind turbines.

[0047] S3: After analyzing the real-time power generation of the wind farm and the real-time power demand of the power grid, generate a wind farm power control strategy.

[0048] The control system obtains wind farm information based on the API interface of the wind farm management platform. The wind farm information includes the number of wind turbines and the location of each turbine. All wind turbines in the wind farm are numbered according to their location. During the operation of the wind farm, the system monitors the operating data of each turbine in real time and calculates the anomaly degree of the turbines through regression analysis algorithm. First, turbines that do not support operation are shut down based on the anomaly degree. Then, the anomaly degree is combined with a clustering algorithm to complete the clustering of the remaining turbines. Clustering the remaining turbines by anomaly degree can group turbines with similar health status into one category, which facilitates subsequent cluster analysis and optimized control. The maximum speed limit of the turbines in each category is corrected.

[0049] The system monitors the real-time rotational speed of wind turbines in each category and dynamically adjusts the blade pitch angle based on the comparison between the real-time rotational speed and the maximum limiting rotational speed, thereby controlling the wind-receiving area of ​​the blades. When there are many wind turbines in a wind farm, the wind turbines located behind the wind direction may experience lower wind speeds due to obstruction from the wind turbines in front. In this case, to ensure power generation efficiency, the blade pitch angle of the rear wind turbines can be increased to bring their real-time rotational speed closer to the maximum limiting rotational speed. When the wind speed increases and causes the real-time rotational speed of the wind turbines to exceed the maximum limiting rotational speed, the blade pitch angle needs to be reduced to reduce the real-time rotational speed to less than or equal to the maximum limiting rotational speed in order to ensure stable operation of the wind turbines. After analyzing the real-time power generation of the wind farm and the real-time power demand of the grid, a wind farm power control strategy is generated.

[0050] This application calculates the anomaly degree of wind turbines using regression analysis algorithms, then combines the anomaly degree with a clustering algorithm to cluster the remaining wind turbines. After clustering the remaining turbines using the anomaly degree, the maximum speed limit of the turbines in each category is corrected, and the real-time speed of the turbines in each category is monitored. Based on the comparison between the real-time speed and the maximum speed limit, the blade pitch angle of the wind turbines is dynamically adjusted. After analyzing the real-time power generation of the wind farm and the real-time power demand of the grid, a wind farm power control strategy is generated. The control system can effectively improve the coordination between wind turbines, making the operation of the wind turbine group more efficient, avoiding the negative impact of individual turbine malfunctions on overall power generation efficiency. By analyzing the real-time power generation of the wind farm and the real-time power demand of the grid, the wind farm power control strategy is generated, improving the collaborative working efficiency between the wind farm and the grid.

[0051] Example 2: The control system obtains wind farm information based on the API interface of the wind farm management platform. The wind farm information includes the number of wind turbines and the location of each turbine. All wind turbines in the wind farm are numbered sequentially according to their location, including the following steps:

[0052] The control system obtains basic information about the wind farm through the API interface provided by the wind farm management platform, including the number of wind turbines and the geographical location information of each turbine (such as longitude, latitude, azimuth, and specific installation location). This information may be returned in the form of a list, containing a unique identifier for each turbine and its corresponding location information. The system calls the API to obtain wind turbine data from the wind farm and parses the returned data format (such as JSON, XML, etc.).

[0053] After obtaining the wind turbine location information, the control system needs to analyze the relative positions of the individual turbines within the wind farm. Typically, the distance between turbines can be calculated using their longitude and latitude to determine their relative positions. For larger wind farms, accuracy can be improved by using a coordinate system (such as the UTM coordinate system). The actual distance between turbines can be calculated using spherical distance formulas (such as the Havesing formula), or a preliminary calculation can be performed using a simplified linear distance formula.

[0054] Based on the relative positions of the wind turbines within a wind farm, a numbering rule should be designed. This rule could be: numbering by the geographical location of the turbines (e.g., starting from a reference point within the wind farm and numbering sequentially by row and column); numbering by the construction order of the wind farm, i.e., turbines built earlier are numbered first; or numbering by the region or wind direction where the turbines are located (for example, the wind farm can be divided into different regions, and the turbines within each region can be numbered sequentially).

[0055] Example of numbering rules: If a wind farm has a regular rectangular layout, it can be divided into several rows and columns, and each wind turbine can be numbered sequentially according to its row and column. The numbering of wind turbines within a wind farm can be divided into multiple areas based on wind direction, with the numbering order depending on wind direction and relative position.

[0056] According to predefined numbering rules, all wind turbines within the wind farm are numbered sequentially. The turbine number may include the following information: wind farm number (used to distinguish different wind farms); the row and column position or area where the turbine is located; and a unique turbine number (e.g., numbered from left to right, from front to back, or based on the wind farm's turbine layout). Integer numbers, letter numbers, or more complex ID systems (such as "area number-turbine number") can be used to uniquely identify each turbine.

[0057] Once the wind turbines are numbered, the control system will generate a list of turbine numbers, recording the number and corresponding location information of each turbine. This list can be used for subsequent turbine status monitoring, operation and maintenance management, and turbine scheduling control. Each number corresponds to the turbine's geographical location information (such as latitude and longitude), health status, power output, fault status, etc. The numbers and related information are stored in a database to ensure quick retrieval and access in subsequent operations.

[0058] During the operation of a wind farm, after real-time monitoring of the operating data of each wind turbine, the anomaly degree of the wind turbine is calculated using a regression analysis algorithm, including the following steps:

[0059] Monitor the wind turbine's operating data, including blade curvature index, generator temperature and noise rise rate, gearbox vibration frequency, and main shaft strain factor;

[0060] The abnormality of the wind turbine is calculated by substituting the blade bending index, generator temperature and noise rise rate, gearbox vibration frequency, and main shaft strain factor into the regression analysis algorithm. The expression is as follows:

[0061] In the formula, unusual represents the degree of abnormality, {y1, y2, y3, y4} represent the blade bending degree index, generator temperature and noise rise rate, gearbox vibration frequency, and main shaft strain factor, respectively, and {w1, w2, w3, w4} represent the regression coefficients of each operating data, and the regression coefficients are greater than 0.

[0062] The logical components of the anomaly degree used in this invention are as follows: taking the impact of operating data on the stability of wind turbine operation as an example, the first is the indicator, that is, the factor that causes the change in the stability of wind turbine operation (in this invention, the impact of operating data on the stability of wind turbine operation); the second is the weight of these indicators, that is, the proportion that each type of operating data occupies when it is generated; the third is the calculation equation, that is, what kind of mathematical operation process is used to obtain the result, and the anomaly degree obtained by calculating the indicators with their respective weights through the calculation equation.

[0063] The operational data obtained from the sample were transformed and processed into a data language recognizable by computer software. Secondly, these evaluation factors were analyzed using SPSS software through Logistic Regression to identify factors and their weights that were significantly correlated with the results. Thirdly, the evaluation factors and weights were substituted into the Logistic Regression equation to obtain the results, specifically:

[0064] First, ensure the integrity of the running data, handling missing and outlier values. Convert the data to a format that SPSS software can recognize, typically storing it as .csv, .xlsx, etc., and then import it into SPSS. Open SPSS, import the processed data file, and transform the variables as needed. For example, for continuous variables, perform standardization or normalization. Select the "Analyze" menu, then select the "Regression" option and choose the "Bivariate Logistic" option. In the dialog box, add the dependent variable (outcome) and independent variables (running data) to the corresponding boxes. SPSS will then fit a Logistic regression model based on the selected variables. In the output results, you will see information such as the model's coefficients, standard errors, and p-values. Examining the coefficients and p-values ​​helps determine which variables are significantly correlated with the results. Generally, a p-value less than 0.05 is considered significant. While fitting the model, variable selection methods, such as stepwise regression, are used to help screen the most relevant factors. Based on the coefficients of the Logistic regression model, the magnitude of the coefficients reflects the degree of influence of each factor on the result, and the sign of the coefficients indicates the direction of influence. After obtaining the significant factors and their coefficients, the Logistic regression equation is obtained. This equation is used to calculate the probability of each sample and thus predict the results.

[0065] The logic for obtaining the generator temperature-noise rise rate is as follows: Obtain the generator's current temperature and noise level (in decibels); obtain the generator's temperature and noise level (in decibels) at the next moment; subtract the next moment's temperature from the current temperature to obtain the temperature rise value; subtract the next moment's noise level from the current noise level to obtain the noise rise value; subtract the previous moment's noise level from the next moment to obtain the monitoring duration; divide the temperature rise value by the monitoring duration to obtain the temperature rise rate; divide the noise rise value by the monitoring duration to obtain the noise rise rate; add the temperature rise rate to the noise rise rate to obtain the generator temperature-noise rise rate. A higher generator temperature-noise rise rate indicates a greater abnormality in generator operation. A high temperature-noise rise rate indicates a high generator workload, insufficient heat dissipation, and potential overheating. This typically leads to overload operation of the wind turbine, thus affecting its operational stability. Sustained high temperatures may accelerate the aging of electrical components and even cause malfunctions.

[0066] The logic for obtaining the gearbox vibration frequency is as follows: during the monitoring period, the number of vibrations of the gearbox during operation is obtained, and the number of vibrations of the gearbox during operation is divided by the monitoring period to obtain the gearbox vibration frequency. A higher vibration frequency may indicate that the gearbox has a fault or abnormal operating state, such as gear damage, insufficient lubrication or installation problems, which leads to unstable operation of the fan, increased noise, or even failure.

[0067] The calculation logic for the blade curvature index is as follows: obtain the number of blades on the wind turbine and calculate the curvature of each blade, expressed as: In the formula, θ(x) represents the degree of blade curvature at the detection point x, E is the elastic modulus of the blade material, I is the moment of inertia of the blade cross section, and M(x) is the bending moment acting at the detection point x. The blade curvature index of the wind turbine is calculated based on the degree of curvature of each blade, and the expression is: In the formula, Let θ(x) be the blade curvature index, n be the number of blades on the wind turbine, and θ(x) be the blade bending index. i Let x be the degree of bending of the i-th blade at the detection point x. The expression for calculating the bending moment acting at the detection point x is: In the formula, M(x) is the bending moment acting at the detection point x, and W is the weight of the blade. Let x represent the distance from the point of application of gravity to the detection point, and L be the length of the blade. The expression for calculating the weight of the blade is: In the formula, W is the weight of the blade, ρ mat Here, A(x) is the density of the blade material, A(x) is the cross-sectional area of ​​the blade at the detection point x, and g is the acceleration due to gravity (9.81 m / s²). 2(L) represents the blade length. A higher blade curvature index usually indicates that the blade is under a heavy load or has structural problems, which may lead to blade fatigue, cracks, or breakage. This can affect the stable operation of the wind turbine, especially at high wind speeds, potentially causing stalling, vibration, or even turbine shutdown.

[0068] The logic for obtaining the principal spindle strain factor is as follows: The strain or deformation of the principal spindle is measured using strain gauges on the spindle, and then the principal spindle strain factor is calculated. The expression is: In the formula, ε is the principal axis strain factor, and V out V is the output voltage of the strain gauge bridge. in G is the input voltage of the strain gauge bridge, and G is the sensitivity factor of the strain gauge. A higher spindle strain factor indicates that the spindle undergoes greater deformation during stress, which may be due to excessive torque, design defects, material aging, or other reasons, leading to spindle fatigue and damage. Excessive spindle deformation can affect the fan's speed control, operating accuracy, and safety, increasing the risk of malfunction.

[0069] Shutting down unsupported fans based on anomaly level includes the following steps:

[0070] The higher the anomaly level of the fan, the more severe the anomaly, and the less support the fan is for operation. The obtained anomaly level is compared with a preset second anomaly threshold. The second anomaly threshold is used to analyze whether the fan can support operation. If the fan's anomaly level is less than or equal to the second anomaly threshold, the fan is analyzed to be normal and can support operation. If the fan's anomaly level is greater than the second anomaly threshold, the fan is analyzed to be abnormal, cannot support operation, and needs to be shut down.

[0071] The remaining wind turbines are clustered using anomaly detection combined with clustering algorithms. Clustering the remaining wind turbines by anomaly detection allows turbines with similar health conditions to be grouped together, facilitating subsequent cluster analysis and optimized control. The process includes the following steps:

[0072] S2.1: Obtain the remaining support turbine numbers and quantities, and generate an initial K value (i.e., the number of clusters) based on the remaining turbine numbers. In the formula, N represents the number of remaining wind turbines. Indicates rounding down;

[0073] S2.2: Randomly calculate the anomaly of K wind turbines as the initial cluster center value, and calculate the difference between the remaining wind turbines and each initial cluster center value. Assign the remaining wind turbines to the cluster with the smallest difference.

[0074] S2.3: After all remaining wind turbines have been allocated, calculate the mean anomaly of each cluster as the new cluster center value;

[0075] S2.4: Repeat steps S2.2 and S2.3. When the maximum number of iterations is reached or the change in the cluster center value is less than the change threshold, it is determined that the convergence condition is met, and K clusters are output.

[0076] The change in cluster center value is less than the change threshold. For example, the difference between the cluster center value of the next iteration and the cluster center value of the previous iteration is less than 0.1.

[0077] The maximum speed limit for fans in each category is calibrated, including the following steps:

[0078] When the abnormality of the fan is less than or equal to the second abnormality threshold, it indicates that the fan can still operate. However, in practical applications, if the abnormality of the fan is less than or equal to the second abnormality threshold, but is very close to the second abnormality threshold, it indicates that the fan's performance has deteriorated. If the fan still maintains its initial maximum speed limit at this time, it may cause the fan to malfunction during operation. To avoid such problems, this application proposes the following solution:

[0079] After obtaining the K clusters and their cluster center values, the maximum speed limit of the fan in each cluster is adjusted based on the cluster center values. The expression is as follows:

[0080] rsp correct =rsp initial *(1-unusual avg In the formula, rsp correct To correct the maximum speed limit, rsp initial The initial maximum speed limit, unusual avg Let be the cluster center value of the cluster, and In the formula, N represents the number of remaining wind turbines, unusual i Let be the anomaly degree of the i-th remaining wind turbine.

[0081] However, in practical applications, some wind turbines may not require maximum speed limit correction. These turbines generally have good overall performance, and correcting their maximum speed limit would reduce the overall power generation efficiency of the wind farm. Therefore, we compare the anomaly degree of the remaining operating wind turbines with a first anomaly threshold. If the first anomaly threshold is less than a second anomaly threshold, the analysis determines whether maximum speed limit correction is needed. If the anomaly degree is less than or equal to the first anomaly threshold, no correction is needed. If the anomaly degree is greater than the first anomaly threshold, correction is required. The maximum speed limit of the wind turbines in each cluster is corrected based on the cluster center value, expressed as: rsp correct =rsp initial *(1-unusual avgIn the formula, rsp correct To correct the maximum speed limit, rsp initial The initial maximum speed limit, unusual avg Let be the cluster center value of the cluster, and In the formula, N represents the number of remaining wind turbines, unusual i Let be the anomaly degree of the i-th remaining wind turbine.

[0082] The real-time rotational speed of the wind turbines in each category is monitored, and the blade pitch angle is dynamically adjusted based on the comparison between the real-time rotational speed and the maximum limiting rotational speed, thereby controlling the wind-receiving area of ​​the blades. This includes the following steps:

[0083] During the operation of the wind farm, the real-time speed of each internal wind turbine is monitored. If the real-time speed of the turbine is less than the corrected maximum speed limit, the pitch angle of the turbine blades is increased until the real-time speed of the turbine equals the corrected maximum speed limit (this allows the turbines located on the rear side of the wind direction to operate in the best operating condition without exceeding the corrected maximum speed limit, which can improve the overall power generation efficiency of the wind farm). If the real-time speed of the turbine is greater than the corrected maximum speed limit, the pitch angle of the turbine blades is decreased until the real-time speed of the turbine equals the corrected maximum speed limit (this can prevent the turbines located on the front side of the wind direction from running too fast, which could lead to failure).

[0084] In practical applications, the adjustment of the real-time speed of the fan may not be so precise, that is, the adjustment is made so that the real-time speed of the fan is equal to the corrected maximum limit speed. Therefore, in this application, when the speed difference obtained by subtracting the real-time speed from the corrected maximum limit speed is between 0 and 0.5 RPM, it is determined that the real-time speed of the fan is equal to the corrected maximum limit speed.

[0085] After analyzing the real-time power generation of the wind farm and the real-time power demand of the power grid, a wind farm power control strategy is generated, including the following steps:

[0086] The wind farm's monitoring system acquires real-time data on wind turbine power generation during operation. This data is then aggregated to determine the wind farm's total power generation. Through a communication interface with the power grid, the system obtains real-time data on the grid's current electricity demand and calculates the difference between the wind farm's real-time power generation and the grid's real-time demand. If the wind farm's power generation exceeds the grid's demand, it may be necessary to store the excess energy or reduce the wind farm's power generation. Conversely, if the wind farm's power generation is less than the grid's demand, it may be necessary to increase power generation or supplement it with backup power.

[0087] Strategies for handling excess power generation:

[0088] Energy storage solutions: If the wind farm generates more electricity than the grid needs, the excess electricity can be stored in an energy storage system (such as a battery energy storage system) for later use.

[0089] Reduced power generation strategy: If there is no energy storage facility, or the energy storage capacity has reached its limit, the wind farm may need to reduce the power generation of the wind turbine by adjusting the pitch angle of the wind turbine to avoid the power surplus from impacting the power grid.

[0090] Priority adjustment: Prioritize the wind turbines in the wind farm, giving priority to those with higher health and better location, and adjust their power generation to ensure overall power generation efficiency.

[0091] Low power generation compensation strategies:

[0092] Backup power activation: If the power generation of the wind farm is insufficient to meet the grid demand, backup power generation equipment (such as gas turbines, energy storage facilities, etc.) can be activated to supplement it.

[0093] Example 3: This example describes a wind farm turbine and a wind farm intelligent optimization control system. The control system includes a numbering module, a clustering module, a dynamic control module, and a strategy generation module.

[0094] Numbering module: Obtains wind farm information based on the API interface of the wind farm management platform, and numbers all wind turbines in the wind farm according to their location. The numbering information is then sent to the clustering module.

[0095] Clustering module: During the operation of the wind farm, after real-time monitoring of the operating data of each wind turbine, the abnormality of the wind turbine is calculated by regression analysis algorithm. First, the wind turbines that do not support operation are shut down based on the abnormality. Then, the abnormality is combined with the clustering algorithm to complete the clustering of the remaining wind turbines. The clustering results are sent to the dynamic control module.

[0096] Dynamic control module: Monitors the real-time speed of wind turbines in each category, and dynamically adjusts the blade pitch angle of the wind turbine based on the comparison between the real-time speed and the maximum speed limit, thereby controlling the wind-receiving area of ​​the blades. The power generation of the wind turbine after the pitch angle adjustment is sent to the strategy generation module.

[0097] Strategy generation module: After analyzing the real-time power generation of the wind farm and the real-time power demand of the power grid, it generates the power control strategy for the wind farm.

[0098] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0099] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0100] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0101] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0102] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A wind turbine and a wind farm intelligent optimization control method, characterized in that: The control method includes the following steps: S1: The control system obtains wind farm information based on the API interface of the wind farm management platform, and numbers all wind turbines in the wind farm according to their location. During the operation of the wind farm, the system monitors the operating data of each wind turbine in real time and calculates the anomaly degree of the wind turbine through regression analysis algorithm. S2: First, shut down the wind turbines that do not support operation based on the anomaly. Then, combine the anomaly with the clustering algorithm to complete the clustering of the remaining wind turbines. Correct the maximum speed limit of the wind turbines in each category and monitor the real-time speed of the wind turbines in each category. Based on the comparison between the real-time speed and the maximum speed limit, dynamically adjust the blade pitch angle of the wind turbines. S3: After analyzing the real-time power generation of the wind farm and the real-time power demand of the power grid, generate a wind farm power control strategy. The maximum speed limit for fans in each category is calibrated, including the following steps: After obtaining the K clusters and their cluster center values, the maximum speed limit of the fan in each cluster is adjusted based on the cluster center values. The expression is as follows: In the formula, To correct the maximum speed limit, The initial maximum speed limit, Let be the cluster center value of the cluster, and In the formula, This represents the number of remaining wind turbines. For the first The abnormality of the remaining wind turbines.

2. The wind turbine and intelligent optimization control method for wind farms according to claim 1, characterized in that: After real-time monitoring of the operating data of each wind turbine, the anomaly degree of the wind turbine is calculated using a regression analysis algorithm, including the following steps: Monitor the wind turbine's operating data, including blade curvature index, generator temperature and noise rise rate, gearbox vibration frequency, and main shaft strain factor; The abnormality of the wind turbine is calculated by substituting the blade bending index, generator temperature and noise rise rate, gearbox vibration frequency, and main shaft strain factor into the regression analysis algorithm.

3. The wind turbine and intelligent optimization control method for a wind farm according to claim 2, characterized in that: The remaining wind turbines are clustered using anomaly detection combined with clustering algorithms, including the following steps: S2.1: Obtain the remaining support turbine numbers and quantities, and generate an initial K value based on the remaining turbine quantity, i.e., the number of clusters. In the formula, N represents the number of remaining wind turbines. Indicates rounding down; S2.2: Randomly calculate the anomaly of K wind turbines as the initial cluster center value, and calculate the difference between the remaining wind turbines and each initial cluster center value. Assign the remaining wind turbines to the cluster with the smallest difference. S2.3: After all remaining wind turbines have been allocated, calculate the mean anomaly of each cluster as the new cluster center value; S2.4: Repeat steps S2.2 and S2.

3. When the maximum number of iterations is reached or the change in the cluster center value is less than the change threshold, it is determined that the convergence condition is met, and K clusters are output.

4. The wind turbine and intelligent optimization control method for a wind farm according to claim 3, characterized in that: The correction of the maximum speed limit for wind turbines in each category also includes the following steps: The anomaly degree of supporting the operation of the remaining wind turbines is compared with the first anomaly threshold. The first anomaly threshold is less than the second anomaly threshold. The first anomaly threshold is used to analyze whether the maximum speed limit of the wind turbines needs to be corrected. If the abnormality of the fan is less than or equal to the first abnormality threshold, the analysis does not require correction of the maximum speed limit of the fan. If the anomaly of a fan exceeds the first anomaly threshold, the analysis requires correction to the maximum speed limit of that fan. This correction is based on the cluster center value, and the expression is as follows: In the formula, To correct the maximum speed limit, The initial maximum speed limit, Let be the cluster center value of the cluster, and In the formula, This represents the number of remaining wind turbines. For the first The abnormality of the remaining wind turbines.

5. The wind turbine and intelligent optimization control method for a wind farm according to claim 4, characterized in that: The dynamic adjustment of the wind turbine blade pitch angle based on the comparison between real-time rotational speed and maximum limiting rotational speed includes the following steps: During the operation of the wind farm, the real-time speed of each internal wind turbine is monitored. If the real-time speed of the wind turbine is less than the corrected maximum limit speed, the pitch angle of the wind turbine blades is increased until the real-time speed of the wind turbine is equal to the corrected maximum limit speed. If the real-time speed of the wind turbine is greater than the corrected maximum speed limit, the blade pitch angle of the wind turbine is reduced until the real-time speed of the wind turbine equals the corrected maximum speed limit.

6. The wind turbine and intelligent optimization control method for a wind farm according to claim 2, characterized in that: Shutting down unsupported fans based on anomaly level includes the following steps: The higher the anomaly level of the fan, the more severe the anomaly, and the less support the fan is for operation. The obtained anomaly level is compared with a preset second anomaly threshold. The second anomaly threshold is used to analyze whether the fan can support operation. If the fan's anomaly level is less than or equal to the second anomaly threshold, the fan is analyzed to be normal and can support operation. If the fan's anomaly level is greater than the second anomaly threshold, the fan is analyzed to be abnormal, cannot support operation, and needs to be shut down.

7. The wind turbine and intelligent optimization control method for a wind farm according to claim 6, characterized in that: The calculation logic for the blade curvature index is as follows: obtain the number of blades on the wind turbine and calculate the curvature of each blade, expressed as: In the formula, Indicates the position of the blade at the detection point. The degree of curvature at that point, It is the elastic modulus of the blade material. It is the moment of inertia of the blade's cross section. It acts at the detection point location. The bending moment at each point is calculated based on the degree of bending of each blade to obtain the blade bending degree index of the wind turbine, expressed as: In the formula, This is an index representing the degree of blade curvature. This refers to the number of blades on the wind turbine. For the first Each leaf at the detection point The degree of curvature at that point.

8. The wind turbine and intelligent optimization control method for a wind farm according to claim 7, characterized in that: Acting at the detection point location The expression for calculating the bending moment at point is: In the formula, It acts at the detection point location. Bending moment at the point, For the weight of the blade, Indicates the position from the point of application of gravity to the detection point. distance, This refers to the length of the blade. The expression for calculating the weight of the blade is: In the formula, For the weight of the blade, It is the density of the blade material. Location of the detection point The cross-sectional area of ​​the blade at that location. It is gravitational acceleration. It refers to the length of the leaf blade.

9. A wind turbine and intelligent optimization control system for a wind farm, used to implement the control method according to any one of claims 1-8, characterized in that: Includes a numbering module, a clustering module, a dynamic control module, and a strategy generation module: Numbering module: Based on the API interface of the wind farm management platform, wind farm information is obtained, and all wind turbines in the wind farm are numbered sequentially according to their location. Clustering module: During the operation of the wind farm, after real-time monitoring of the operating data of each wind turbine, the anomaly degree of the wind turbine is calculated by regression analysis algorithm. First, the wind turbines that do not support operation are shut down based on the anomaly degree. Then, the anomaly degree is combined with the clustering algorithm to complete the clustering process of the remaining wind turbines. Dynamic control module: Monitors the real-time speed of the wind turbines in each category, and dynamically adjusts the blade pitch angle based on the comparison between the real-time speed and the maximum limit speed, thereby controlling the wind-receiving area of ​​the blades; Strategy generation module: After analyzing the real-time power generation of the wind farm and the real-time power demand of the power grid, it generates the power control strategy for the wind farm.

Images