A wind turbine yaw attitude control system and method

Abstract

The present invention belongs to the technical field of wind turbines, and relates to a yaw attitude control system and method for a wind turbine, comprising: a wind condition acquisition module: acquiring historical operating environment data of the wind turbine, and normalizing the historical wind condition data of the wind turbine through feature extraction, and dividing the data into a training set and a test set; a wind condition prediction module: constructing a wind condition prediction model, iteratively training the wind condition prediction model with the training set until the iteration is completed, evaluating the wind condition prediction model with the test set to obtain a trained wind condition prediction model, and predicting the wind condition prediction data based on the real-time operating environment data of the wind turbine through the wind condition prediction model to obtain wind condition prediction data; a yaw control module: executing the yaw system control logic of the wind turbine based on the current wind condition data and the input wind condition prediction data and combined with the maximum value of the wind condition data, adjusting the yaw system attitude of the wind turbine, and stopping yaw after the wind adjustment is completed.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbines, and in particular to a yaw attitude control system and method for a wind turbine set. Background Art

[0002] The yaw system of a wind turbine primarily controls the direction of the rotor to ensure the turbine faces the wind, thereby maximizing wind energy capture. Currently, due to the high uncertainty of wind direction caused by variable wind conditions, the yaw system often needs to be activated and deactivated based on real-time wind data. This can affect the stable operation of the wind turbine and can also damage the system over time, resulting in significant economic losses. To improve the operational stability of wind turbines, we have designed a yaw attitude control system and method for wind turbines. Summary of the Invention

[0003] The object of the present invention is to provide a wind turbine yaw attitude control system and method, which combines the BP model and the PSO algorithm to construct a wind condition prediction model, which can effectively predict the wind conditions in the next time period, and combines the preset yaw system control logic to comprehensively regulate the yaw system according to the average value of each time period, effectively overcoming the damage or failure that is easily caused by frequent starting and stopping of the yaw system for a long time, and improving the stability of the wind turbine operation.

[0004] The embodiments of the present invention are achieved through the following technical solutions:

[0005] A wind turbine yaw attitude control system, comprising:

[0006] Wind condition acquisition module: This module obtains the historical operating environment data of the wind turbine and obtains the historical wind condition data of the wind turbine through feature extraction. After normalization, the historical wind condition data of the wind turbine is divided into a training set and a test set.

[0007] Wind condition prediction module: Build a wind condition prediction model, iteratively train the wind condition prediction model with the training set, evaluate the wind condition prediction model with the test set until the iteration is complete, and obtain a trained wind condition prediction model. The wind condition prediction model is used to make predictions based on the real-time operating environment data of the wind turbine to obtain wind condition prediction data.

[0008] Yaw control module: Based on the current wind condition data and the input wind condition forecast data, combined with the maximum value of the wind condition data, the yaw system control logic of the wind turbine is executed to adjust the yaw system attitude of the wind turbine until the wind is aligned and the yaw is stopped.

[0009] Optionally, a pre-processing sub-module is preset in the wind condition acquisition module, and the pre-processing sub-module is used to pre-process the historical operating environment data of the wind turbine generator set to perform feature extraction.

[0010] Optionally, the historical wind condition data of the wind turbine generator set is normalized, and the specific calculation formula of the normalization process is:

[0011]

[0012] Among them, Q i is the historical wind condition data of the wind turbine, Q i ' is the normalized historical wind data, Q max is the maximum value of historical wind data before normalization, Q min is the minimum value of historical wind data before normalization.

[0013] Optionally, the specific training process of the wind condition prediction model is as follows:

[0014] The wind condition prediction model is constructed by combining the BP neural network model and the PSO algorithm.

[0015] Initialize the PSO parameters of the wind prediction model, including the number of PSOs, the maximum number of iterations, the local learning factor, and the global learning factor. Randomly obtain the initial position and speed of the PSO within the initialization value range. At the same time, construct the fitness function during the wind prediction model training process and iteratively update the PSO parameters through the fitness function.

[0016] Determine whether the iteration stops. If not, continue iterating until the maximum number of iterations is reached. If so, determine whether the number of iterations reaches the maximum number of iterations. If not, return to the previous step until the maximum number of iterations is reached. If so, complete the forward propagation of the wind prediction model and update the parameters of the wind prediction model through gradient backpropagation.

[0017] After the back propagation is completed, the trained wind condition prediction model is output, and the trained wind condition prediction model is evaluated based on the test set, and the trained wind condition prediction model is output based on the evaluation result.

[0018] Optionally, the fitness function in the wind condition prediction model training process is constructed, and the specific calculation formula of the fitness function is as follows:

[0019]

[0020] Among them, E ak is the connection weight between the input layer and the hidden layer, a and k are both constants, N k is the threshold of the hidden layer, M a is the feature matrix.

[0021] Optionally, the wind condition data specifically refers to wind direction and wind speed data.

[0022] Optionally, a first set value A, a second set value B, and a third set value C of wind speed data are set in the yaw system control logic of the wind turbine generator set, wherein the first set value A is specifically a minimum value, the second set value B is specifically a standard value, and the third set value C is specifically a maximum value. The specific steps include:

[0023] Setting a first time period based on current wind condition data, solving an average value of current wind direction data within the first time period, and calculating a deviation angle between the average value of the current wind direction data and a nacelle of the wind turbine generator set;

[0024] Based on the absolute value of the deviation angle, it is determined whether it is lower than the minimum yaw angle of the yaw system. If it is lower, the yaw system is not activated; if it is greater than or equal to it, the process proceeds to the next step.

[0025] The current wind condition data is predicted using the trained wind condition prediction model to obtain the wind direction and wind speed data for the second time period;

[0026] Determine whether the wind speed data is less than a first set value A. If so, stop starting the yaw system of the wind turbine generator set; if not, determine whether the wind speed data is between the first set value A and the second set value B. If so, start the yaw system of the wind turbine generator set, and control the yaw system of the wind turbine generator set to yaw at a preset first yaw speed until the wind speed data is lower than the first set value A, and stop the yaw system of the wind turbine generator set; if not, determine whether the wind speed data is between the second set value B and the third set value C. If not, determine that the wind speed data is greater than the third set value C, and stop starting the yaw system of the wind turbine generator set. If so, start the yaw system of the wind turbine generator set, and control the yaw system of the wind turbine generator set to yaw at the preset second yaw speed until the wind speed data is lower than the first set value A, and stop the yaw system of the wind turbine generator set.

[0027] A method for controlling the yaw attitude of a wind turbine generator system, the method comprising the following steps:

[0028] Obtain historical operating environment data of the wind turbine and obtain historical wind condition data of the wind turbine through feature extraction. After normalization, the historical wind condition data of the wind turbine is divided into a training set and a test set;

[0029] Construct a wind condition prediction model, iteratively train the wind condition prediction model with the training set until the iteration is complete, evaluate the wind condition prediction model with the test set, and obtain a trained wind condition prediction model. Use the wind condition prediction model to make predictions based on the real-time operating environment data of the wind turbine to obtain wind condition prediction data.

[0030] Based on the current wind condition data and the input wind condition forecast data, combined with the maximum value of the wind condition data, the yaw system control logic of the wind turbine is executed to adjust the yaw system attitude of the wind turbine until the wind is aligned and the yaw is stopped.

[0031] The technical solutions of the embodiments of the present invention have at least the following advantages and beneficial effects:

[0032] The embodiment of the present invention combines the BP model and the PSO algorithm to construct a wind condition prediction model, which can effectively predict the wind conditions in the next time period. It also combines the preset yaw system control logic to comprehensively regulate the yaw system according to the average value of each time period, effectively overcoming the damage or failure that may be easily caused by frequent starting and stopping of the yaw system for a long time, and improving the stability of wind turbine operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of the principle of a wind turbine yaw attitude control system provided by an embodiment of the present invention;

[0034] Figure 2 A schematic flow chart of a method for controlling the yaw attitude of a wind turbine generator system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0036] like Figure 1 As shown, the present invention provides one embodiment: a wind turbine yaw attitude control system, comprising:

[0037] Wind condition acquisition module: This module obtains the historical operating environment data of the wind turbine and obtains the historical wind condition data of the wind turbine through feature extraction. After normalization, the historical wind condition data of the wind turbine is divided into a training set and a test set.

[0038] Wind condition prediction module: Build a wind condition prediction model, iteratively train the wind condition prediction model with the training set, evaluate the wind condition prediction model with the test set until the iteration is complete, and obtain a trained wind condition prediction model. The wind condition prediction model is used to make predictions based on the real-time operating environment data of the wind turbine to obtain wind condition prediction data.

[0039] Yaw control module: Based on the current wind condition data and the input wind condition forecast data, combined with the maximum value of the wind condition data, the yaw system control logic of the wind turbine is executed to adjust the yaw system attitude of the wind turbine until the wind is aligned and the yaw is stopped.

[0040] In this embodiment, a pre-processing sub-module is preset in the wind condition acquisition module, and the pre-processing sub-module is used to pre-process the historical operating environment data of the wind turbine generator set so as to perform feature extraction.

[0041] It can be understood that the data preprocessing preset in the preprocessing submodule includes steps such as data cleaning and abnormal data removal, which are used to process the historical data of the wind turbine generator set so as to extract the historical wind condition data of the wind turbine generator set.

[0042] Specifically, the historical wind condition data of the wind turbine generator set is normalized, and the specific calculation formula of the normalization process is:

[0043]

[0044] Among them, Q i is the historical wind condition data of the wind turbine, Q i ' is the normalized historical wind data, Q max is the maximum value of historical wind data before normalization, Q min is the minimum value of historical wind data before normalization.

[0045] In this embodiment, the specific training process of the wind condition prediction model is as follows:

[0046] The wind condition prediction model is constructed by combining the BP neural network model and the PSO (particle swarm optimization) algorithm.

[0047] Initialize the PSO parameters of the wind prediction model, including the number of PSOs, the maximum number of iterations, the local learning factor, and the global learning factor. Randomly obtain the initial position and speed of the PSO within the initialization value range. At the same time, construct the fitness function during the wind prediction model training process and iteratively update the PSO parameters through the fitness function.

[0048] Determine whether the iteration stops. If not, continue iterating until the maximum number of iterations is reached. If so, determine whether the number of iterations reaches the maximum number of iterations. If not, return to the previous step until the maximum number of iterations is reached. If so, complete the forward propagation of the wind prediction model and update the parameters of the wind prediction model through gradient backpropagation.

[0049] After the back propagation is completed, the trained wind condition prediction model is output, and the trained wind condition prediction model is evaluated based on the test set, and the trained wind condition prediction model is output based on the evaluation result.

[0050] Specifically, the fitness function in the wind condition prediction model training process is constructed, and the specific calculation formula of the fitness function is as follows:

[0051]

[0052] Among them, E ak is the connection weight between the input layer and the hidden layer, a and k are both constants, N k is the threshold of the hidden layer, M a is the feature matrix.

[0053] More specifically, the wind condition data is wind direction and wind speed data.

[0054] In this embodiment, a first set value A, a second set value B, and a third set value C of wind speed data are set in the yaw system control logic of the wind turbine generator set, wherein the first set value A is specifically a minimum value, the second set value B is specifically a standard value, and the third set value C is specifically a maximum value. The specific steps include:

[0055] Setting a first time period based on current wind condition data, solving an average value of current wind direction data within the first time period, and calculating a deviation angle between the average value of the current wind direction data and a nacelle of the wind turbine generator set;

[0056] Based on the absolute value of the deviation angle, it is determined whether it is lower than the minimum yaw angle of the yaw system. If it is lower, the yaw system is not activated; if it is greater than or equal to it, the process proceeds to the next step.

[0057] The current wind condition data is predicted using the trained wind condition prediction model to obtain the wind direction and wind speed data for the second time period;

[0058] Determine whether the wind speed data is less than a first set value A. If so, stop starting the yaw system of the wind turbine generator set; if not, determine whether the wind speed data is between the first set value A and the second set value B. If so, start the yaw system of the wind turbine generator set, and control the yaw system of the wind turbine generator set to yaw at a preset first yaw speed until the wind speed data is lower than the first set value A, and stop the yaw system of the wind turbine generator set; if not, determine whether the wind speed data is between the second set value B and the third set value C. If not, determine that the wind speed data is greater than the third set value C, and stop starting the yaw system of the wind turbine generator set. If so, start the yaw system of the wind turbine generator set, and control the yaw system of the wind turbine generator set to yaw at the preset second yaw speed until the wind speed data is lower than the first set value A, and stop the yaw system of the wind turbine generator set.

[0059] In the above application, the first time period set in this embodiment can be understood as the maximum yaw time, and the subsequent second time period, third time period, ..., Nth time period are all consistent with the length of the first time period. The first set value A, the second set value B, and the third set value C of the wind speed can be understood as the first set value A corresponding to the yaw system startup wind speed, the second set value B corresponding to the yaw system standard wind speed, and the third set value C corresponding to the yaw system wind speed threshold. The above logic is set based on this. Specifically, the first set value A, the second set value B, and the third set value C are exemplified as: the first set value A is specifically 2.5 meters per second, the second set value B is specifically 6 meters per second, and the third set value C is specifically 20 meters per second. The wind direction at least reaches the minimum yaw angle of 3 degrees of the yaw system.

[0060] like Figure 2 As shown, the present invention also provides another embodiment: a method for controlling the yaw attitude of a wind turbine generator set, the method comprising the following steps:

[0061] Obtain historical operating environment data of the wind turbine and obtain historical wind condition data of the wind turbine through feature extraction. After normalization, the historical wind condition data of the wind turbine is divided into a training set and a test set;

[0062] Construct a wind condition prediction model, iteratively train the wind condition prediction model with the training set until the iteration is complete, evaluate the wind condition prediction model with the test set, and obtain a trained wind condition prediction model. Use the wind condition prediction model to make predictions based on the real-time operating environment data of the wind turbine to obtain wind condition prediction data.

[0063] Based on the current wind condition data and the input wind condition forecast data, combined with the maximum value of the wind condition data, the yaw system control logic of the wind turbine is executed to adjust the yaw system attitude of the wind turbine until the wind is aligned and the yaw is stopped.

[0064] It can be understood that the wind turbine yaw attitude control method provided in this embodiment and the wind turbine yaw attitude control system provided in the above embodiment are based on the same inventive concept. For more specific working principles of each module in the embodiment of the present invention, please refer to the above implementation method and will not be repeated in the embodiment of the present invention.

[0065] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A wind turbine yaw attitude control system, characterized in that: include: Wind condition acquisition module: This module obtains the historical operating environment data of the wind turbine and obtains the historical wind condition data of the wind turbine through feature extraction. After normalization, the historical wind condition data of the wind turbine is divided into a training set and a test set. Wind condition prediction module: Build a wind condition prediction model, iteratively train the wind condition prediction model with the training set, evaluate the wind condition prediction model with the test set until the iteration is complete, and obtain a trained wind condition prediction model. The wind condition prediction model is used to make predictions based on the real-time operating environment data of the wind turbine to obtain wind condition prediction data. Yaw control module: Based on the current wind condition data and the input wind condition forecast data, combined with the maximum wind condition data, it executes the yaw system control logic of the wind turbine and adjusts the yaw system attitude of the wind turbine until the wind is aligned and the yaw is stopped; The normalization process is performed on the historical wind condition data of the wind turbine generator set. The specific calculation formula of the normalization process is: in, is the historical wind condition data of the wind turbine, is the normalized historical wind data, is the maximum value of historical wind data before normalization, is the minimum value of historical wind data before normalization; The specific training process of the wind condition prediction model is as follows: The wind condition prediction model is constructed by combining the BP neural network model and the PSO algorithm. Initialize the PSO parameters of the wind prediction model, including the number of PSOs, the maximum number of iterations, the local learning factor, and the global learning factor. Randomly obtain the initial position and speed of the PSO within the initialization value range. At the same time, construct the fitness function during the wind prediction model training process and iteratively update the PSO parameters through the fitness function. Determine whether the iteration stops. If not, continue iterating until the maximum number of iterations is reached. If so, determine whether the number of iterations reaches the maximum number of iterations. If not, return to the previous step until the maximum number of iterations is reached. If so, complete the forward propagation of the wind prediction model and update the parameters of the wind prediction model through gradient backpropagation. After the back propagation is completed, the trained wind condition prediction model is output, and the trained wind condition prediction model is evaluated based on the test set, and the trained wind condition prediction model is output based on the evaluation results; The fitness function in the wind condition prediction model training process is constructed, and the specific calculation formula of the fitness function is as follows: in, is the connection weight between the input layer and the hidden layer, 、 are all constants, is the threshold of the hidden layer, is the feature matrix.

2. The wind turbine yaw attitude control system according to claim 1, characterized in that: The wind condition acquisition module is preset with a pre-processing sub-module, and the pre-processing sub-module is used to pre-process the historical operating environment data of the wind turbine generator set so as to perform feature extraction.

3. The wind turbine yaw attitude control system according to claim 1 or 2, characterized in that: The wind condition data specifically refers to wind direction and wind speed data.

4. The wind turbine yaw attitude control system according to claim 3, characterized in that: The yaw system control logic of the wind turbine generator set is set with a first set value A, a second set value B, and a third set value C of wind speed data, wherein the first set value A is specifically a minimum value, the second set value B is specifically a standard value, and the third set value C is specifically a maximum value. The specific steps include: Setting a first time period based on current wind condition data, solving an average value of current wind direction data within the first time period, and calculating a deviation angle between the average value of the current wind direction data and a nacelle of the wind turbine generator set; Based on the absolute value of the deviation angle, it is determined whether it is lower than the minimum yaw angle of the yaw system. If it is lower, the yaw system is not activated; if it is greater than or equal to it, the process proceeds to the next step. The current wind condition data is predicted using the trained wind condition prediction model to obtain the wind direction and wind speed data for the second time period; Determine whether the wind speed data is less than a first set value A. If so, stop starting the yaw system of the wind turbine generator set; if not, determine whether the wind speed data is between the first set value A and the second set value B. If so, start the yaw system of the wind turbine generator set, and control the yaw system of the wind turbine generator set to yaw at a preset first yaw speed until the wind speed data is lower than the first set value A, and stop the yaw system of the wind turbine generator set; if not, determine whether the wind speed data is between the second set value B and the third set value C. If not, determine that the wind speed data is greater than the third set value C, and stop starting the yaw system of the wind turbine generator set. If so, start the yaw system of the wind turbine generator set, and control the yaw system of the wind turbine generator set to yaw at the preset second yaw speed until the wind speed data is lower than the first set value A, and stop the yaw system of the wind turbine generator set.

5. A method for controlling the yaw attitude of a wind turbine generator set, applied to the yaw attitude control system of a wind turbine generator set according to any one of claims 1 to 4, characterized in that: The steps of the method include: Obtain historical operating environment data of the wind turbine and obtain historical wind condition data of the wind turbine through feature extraction. After normalization, the historical wind condition data of the wind turbine is divided into a training set and a test set; Construct a wind condition prediction model, iteratively train the wind condition prediction model with the training set until the iteration is complete, evaluate the wind condition prediction model with the test set, and obtain a trained wind condition prediction model. Use the wind condition prediction model to make predictions based on the real-time operating environment data of the wind turbine to obtain wind condition prediction data. Based on the current wind condition data and the input wind condition forecast data, combined with the maximum value of the wind condition data, the yaw system control logic of the wind turbine is executed to adjust the yaw system attitude of the wind turbine until the wind is aligned and the yaw is stopped.

Images