Wind turbine control method and system based on meteorological information

Abstract

The application provides a kind of wind turbine control method and system based on weather information.The steps are as follows: real-time acquisition of multi-dimensional data containing basic load associated data and comprehensive weather parameters, standardization processing and identification of wave excitation dominant frequency and energy concentration band;establish a multi-parameter coupling model of weather multi-parameter and key component dynamic response, output the quantitative parameters of cooperative stress state;based on the parameters, construct a risk assessment index system combining wind direction-wave direction difference, calculate the load asymmetry risk level of each component;according to the risk level, optimize the cooperative adjustment parameters of yaw speed and variable pitch rate with the goal of overall optimization of power generation efficiency, structural safety and fatigue life;execute control and dynamically correct model weight based on operation feedback.Solve the problem that the existing wind turbine control method has insufficient data dimension and inaccurate risk assessment model, resulting in poor adaptability of control strategy and difficulty in global optimization under complex weather conditions.

Description

Technical Field

[0001] This invention relates to the field of wind turbine operation control technology, and more specifically, to a wind turbine control method and system based on meteorological information. Background Technology

[0002] Offshore wind power, as one of the core development directions in the field of clean energy, has long been constrained by the complex environmental loads of the ocean in terms of operational stability and reliability. In extreme weather scenarios such as typhoons and severe storms, wind loads are prone to rapid and significant changes due to the randomness of atmospheric motion, while wave loads are affected by the inertia of water, and their directional changes lag significantly behind wind direction changes. This instantaneous mismatch between wind and wave loads can cause the wind turbine support structure to bear severe asymmetrical loads, which not only aggravates fatigue damage to the foundation structure but may also lead to premature failure of critical components such as blades and nacelles, seriously threatening the safe operation of wind turbines. To address this issue, patent document CN120798659A (a control method, device, terminal, and storage medium for a wind turbine generator) proposes a control scheme that involves real-time synchronous acquisition of wind and wave direction data, calculation of wind direction change rate, and spectral analysis of vibration signals from the supporting structure to identify the dominant frequency components of wave excitation. This is combined with the tower top motion trajectory to determine the concentrated frequency band of wave spectrum energy, thereby assessing the risk of load asymmetry and adaptively adjusting yaw speed and pitch rate. This scheme effectively achieves accurate identification and early warning of wind and wave load direction mismatch conditions, playing a significant role in improving the operational reliability of wind turbine generators under extreme weather conditions.

[0003] However, the above control methods still have significant technical limitations and are difficult to meet the higher requirements for efficient and safe operation of wind turbines under complex weather conditions:

[0004] First, the core data support of existing technologies is limited to local information such as wind direction, wave direction, vibration of supporting structures, and tower top movement. It does not fully integrate more comprehensive meteorological parameters such as wind speed and turbulence intensity, atmospheric pressure fluctuations, precipitation intensity, and temperature gradient. These meteorological factors not only directly affect the wind energy capture efficiency of the wind turbine, but also indirectly affect the distribution pattern and dynamic response characteristics of wind and wave loads by changing airflow viscosity and turbulence characteristics, resulting in insufficient adaptability of existing control strategies to complex and variable meteorological conditions.

[0005] Secondly, the existing risk assessment system mainly relies on the two-dimensional correlation between the difference in wind and wave direction and the frequency band of wave energy concentration. It has not established a coupled model of multiple meteorological parameters and the dynamic response of key components (blades, nacelles, foundations) of wind turbines. It cannot accurately quantify the coordinated stress state of each component under different meteorological combinations, which means that the adjustment of control parameters can only focus on the single dimension of yaw and pitch rate optimization. It is difficult to achieve a global optimal balance between power generation efficiency, structural safety and component fatigue life. Especially in the scenario of sudden changes in meteorological parameters and multiple meteorological disasters, there is still a lot of room for improvement in control accuracy and robustness. Summary of the Invention

[0006] The purpose of this invention is to provide a wind turbine control method and system based on meteorological information, aiming to solve the problems of poor adaptability of control strategies and difficulty in global optimization under complex meteorological conditions caused by insufficient data dimensions and inaccurate risk assessment models in existing wind turbine control methods.

[0007] This invention is achieved through the following technical solution:

[0008] A wind turbine control method based on meteorological information includes the following steps:

[0009] Real-time synchronous acquisition of multi-dimensional data, including basic load correlation data and comprehensive meteorological parameters; wherein, the basic load correlation data includes wind direction, wave direction, vibration signal of support structure and movement trajectory of tower top; the comprehensive meteorological parameters include wind speed turbulence intensity, atmospheric pressure fluctuation, precipitation intensity and temperature gradient;

[0010] Multidimensional data is standardized to obtain a standardized dataset. Spectral analysis is then performed on the standardized vibration signal of the support structure to identify the dominant frequency components of wave excitation. The concentrated frequency band of wave spectrum energy is determined based on the standardized tower top motion trajectory.

[0011] Based on a standardized dataset, a multi-parameter coupling model of meteorological multi-parameters and dynamic response of key components of wind turbine is established. The multi-parameter coupling model takes standardized basic load correlation data and comprehensive meteorological parameters as input and outputs the quantitative parameters of the coordinated stress state of each key component.

[0012] Based on the quantitative parameters of the coordinated stress state, combined with the dominant frequency components of wave excitation, the concentrated frequency band of wave spectrum energy, and the instantaneous difference between wind direction and wave direction, a multi-dimensional risk assessment index system is constructed to calculate the load asymmetry risk level of each key component.

[0013] Based on the load asymmetry risk level, and with the goal of achieving the global optimum of power generation efficiency, structural safety, and component fatigue life, the coordinated adjustment parameters for yaw speed and pitch rate are optimized and generated.

[0014] The coordinated adjustment parameters are sent to the wind turbine actuator for execution control, and the wind turbine operating status feedback data is collected in real time and sent back to the multi-parameter coupled model to dynamically correct the model parameter weights.

[0015] Optionally, the specific process of real-time synchronous acquisition of multi-dimensional data is as follows:

[0016] By deploying sensor arrays in the wind turbine and its marine environment, basic load correlation data and comprehensive meteorological parameters are collected simultaneously;

[0017] The acquisition of the basic load-related data includes: obtaining wind direction and wave direction through wind direction sensor and wave direction sensor respectively; acquiring vibration signals of the support structure through acceleration sensor installed on the support structure; and recording the movement trajectory of the tower top through tower top positioning system.

[0018] The collection of comprehensive meteorological parameters includes: obtaining wind speed and calculating its turbulence intensity through an anemometer; monitoring atmospheric pressure fluctuations through a barometer; obtaining precipitation intensity through a rain gauge; and measuring temperature at different altitudes and calculating temperature gradients through a temperature sensor array.

[0019] The synchronously collected basic load correlation data and comprehensive meteorological parameters are integrated into multi-dimensional data with time-series labels.

[0020] Optionally, the specific process of establishing a multi-parameter coupled model of meteorological multi-parameters and the dynamic response of key components of wind turbines based on a standardized dataset is as follows:

[0021] The standardized dataset is used as the model input, which includes a standardized subset of basic load correlation data and a subset of comprehensive meteorological parameters.

[0022] The model input is subjected to nonlinear transformation and feature fusion through at least one hidden layer, which is used to extract high-dimensional coupling features between meteorological parameters and key component responses;

[0023] Based on the fusion features processed by the hidden layer, the output layer calculates and outputs the quantification parameters of the collaborative stress state of each key component, including blades, nacelle and foundation. The quantification parameters of the collaborative stress state are vectors characterizing the dynamic mechanical response of each component.

[0024] The multi-parameter coupled model is trained using a historical operational data sample set. The weight parameters of each connection layer in the model are iteratively optimized by calculating the output of forward propagation and correcting the error of backward propagation, until the error between the model output of the quantified parameter of the cooperative force state and the true value of the sample is lower than a preset threshold.

[0025] Optionally, the specific process of constructing a multi-dimensional risk assessment index system and calculating the load asymmetry risk level of each key component is as follows:

[0026] Based on the aforementioned quantification parameters of the coordinated stress state, a load asymmetry evaluation factor is defined to characterize the degree of dynamic load imbalance of key components.

[0027] Based on the dominant frequency component of wave excitation, the concentrated frequency band of wave spectrum energy, and the instantaneous difference between wind direction and wave direction, wave excitation frequency evaluation factor, wave energy concentration evaluation factor, and wind-wave direction mismatch evaluation factor are defined respectively.

[0028] The load asymmetry assessment factor, the wave excitation frequency assessment factor, the wave energy concentration assessment factor, and the wind and wave direction mismatch assessment factor are used as the four core assessment factors of the multi-dimensional risk assessment index system. Weight coefficients are assigned to the four core assessment factors, and the weight coefficients are determined based on the correlation analysis between historical meteorological data and component failure data.

[0029] Based on the assigned weight coefficients, the four core assessment factors are weighted and fused to obtain the comprehensive risk assessment value of each key component;

[0030] Based on the preset risk level threshold range, the comprehensive risk assessment value is mapped to the corresponding load asymmetry risk level.

[0031] Optionally, after the multi-dimensional risk assessment index system calculates the load asymmetry risk level of each key component, it further includes the following steps:

[0032] When the load asymmetry risk level of any critical component exceeds the preset severe risk threshold, a simulation sequence of the evolution of meteorological parameters towards extreme working conditions in future time periods is generated based on the standardized dataset and the multi-parameter coupling model at the current moment.

[0033] The simulation sequence is input into the multi-parameter coupled model to calculate the predictive collaborative stress state quantification parameters of each key component in the future time period;

[0034] Based on the predictive collaborative stress state quantification parameters, with loss prevention and power generation maintenance as the core optimization objectives, the preventive collaborative adjustment parameter sequence for the future period is generated in advance through iterative optimization.

[0035] The preventive coordinated adjustment parameter sequence is stored in the control queue. When the actual meteorological conditions change to be close to the simulation sequence, the preventive coordinated adjustment parameters in the control queue are called in sequence and sent to the wind turbine actuator to perform advanced control.

[0036] Optionally, the specific process of optimizing the coordinated adjustment parameters of yaw speed and pitch rate based on the load asymmetry risk level, with the goal of achieving global optimization of power generation efficiency, structural safety, and component fatigue life, is as follows:

[0037] Based on the load asymmetry risk level, the basic yaw speed adjustment coefficient and the basic pitch rate adjustment coefficient are queried and determined from the preset risk level-adjustment strategy mapping table.

[0038] The basic yaw speed adjustment coefficient and the basic pitch rate adjustment coefficient are fused with the quantified parameters of the cooperative stress state output by the multi-parameter coupling model and input to the preset multi-objective optimization controller; the multi-objective optimization controller has a built-in optimization function with the sub-objectives of maximizing power generation efficiency, minimizing structural safety risk, and minimizing component fatigue damage;

[0039] Based on the quantized parameters of the cooperative stress state, the multi-objective optimization controller calculates, under the current stress state, the contribution of adjusting the yaw speed to power generation efficiency, the suppression coefficient of structural safety risk, and the mitigation weight of adjusting the pitch rate for fatigue damage.

[0040] Based on the contribution, suppression coefficient and mitigation weight, the basic yaw speed adjustment coefficient and basic pitch rate adjustment coefficient are weighted and corrected to generate preliminary coordinated adjustment parameters.

[0041] Using the initial coordinated adjustment parameters as the initial solution, a multi-objective optimization algorithm is called to iteratively solve the optimization function. Under the condition of satisfying the dynamic load constraints of each key component, the final yaw speed command value and the final pitch rate command value that achieve the global optimal balance between power generation efficiency, structural safety and component fatigue life are obtained.

[0042] The final yaw speed command value and the final pitch rate command value are integrated into the coordinated adjustment parameters.

[0043] Optionally, the specific process of real-time acquisition of wind turbine operating status feedback data and transmission back to the multi-parameter coupled model, and dynamic correction of model parameter weights, is as follows:

[0044] The operational status feedback data includes real-time stress data of key components, wind turbine speed data, yaw actuator response speed data, and pitch actuator response angle data.

[0045] The deviation between the quantified parameters of the coordinated stress state output by the multi-parameter coupled model and the real-time stress data of key components in the operation status feedback data is calculated. If the deviation exceeds the preset deviation threshold, the model parameter weight correction process is triggered.

[0046] Based on the magnitude and direction of the deviation, the gradient descent method is used to dynamically adjust the connection weights of the basic load associated data input layer, the comprehensive meteorological parameter input layer and the hidden layer in the multi-parameter coupled model. During the adjustment process, the weight parameters are iteratively updated with the goal of minimizing the deviation until the deviation is lower than the preset deviation threshold.

[0047] The adjusted weight parameters are saved to the model parameter library as the latest weight benchmark for subsequent multi-parameter coupled model calculations, and the time series information and corresponding meteorological conditions of this correction are recorded.

[0048] Based on the same inventive concept, the present invention also provides a wind turbine control system based on meteorological information, used to implement the aforementioned wind turbine control method based on meteorological information, comprising:

[0049] The data synchronization acquisition module is used to collect multi-dimensional data in real time through a sensor array deployed on the wind turbine and its marine environment. The multi-dimensional data includes basic load correlation data and comprehensive meteorological parameters. The basic load correlation data includes wind direction, wave direction, vibration signal of the support structure, and movement trajectory of the tower top. The comprehensive meteorological parameters include wind speed turbulence intensity, atmospheric pressure fluctuation, precipitation intensity, and temperature gradient.

[0050] The data processing and analysis module is connected to the data synchronization acquisition module. It is used to standardize the received multi-dimensional data to obtain a standardized dataset. It also performs spectrum analysis based on the standardized support structure vibration signal to identify the dominant frequency component of wave excitation, and determines the concentrated frequency band of wave spectrum energy based on the standardized tower top motion trajectory.

[0051] The model building and calculation module is connected to the data processing and analysis module. It is used to build and run a multi-parameter coupled model of meteorological multi-parameters and dynamic response of key components of wind turbine based on the standardized dataset. The model takes standardized basic load correlation data and comprehensive meteorological parameters as input, and outputs the quantified parameters of the coordinated stress state of each key component after nonlinear transformation and feature fusion through at least one hidden layer.

[0052] The risk assessment module is connected to the data processing and analysis module and the model building and calculation module, respectively. It is used to construct a multi-dimensional risk assessment index system based on the quantitative parameters of the coordinated stress state, combined with the dominant frequency components of wave excitation, the concentrated frequency band of wave spectrum energy and the instantaneous difference between wind direction and wave direction, and to calculate the load asymmetry risk level of each key component.

[0053] The collaborative control parameter optimization module is connected to the risk assessment module and the model construction and calculation module; it is used to optimize and generate collaborative adjustment parameters for yaw speed and pitch rate by calling a multi-objective optimization algorithm based on the load asymmetry risk level and with the global optimization of power generation efficiency, structural safety and component fatigue life as the objectives.

[0054] The execution control and feedback module is connected to the collaborative control parameter optimization module; it is used to send the collaborative adjustment parameters to the yaw and pitch actuators of the wind turbine to perform control; and to collect the operating status feedback data of the wind turbine in real time and send it back to the construction and calculation module to dynamically correct the parameter weights of the multi-parameter coupling model.

[0055] Based on the same inventive concept, the present invention also provides an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the above-described wind turbine control method based on meteorological information.

[0056] Based on the same inventive concept, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described wind turbine control method based on meteorological information.

[0057] The technical solution of the present invention has at least the following advantages and beneficial effects:

[0058] By introducing comprehensive meteorological parameters such as wind speed turbulence intensity, atmospheric pressure fluctuations, precipitation intensity, and temperature gradient, the shortcomings of existing technologies that rely solely on wind direction and wave direction information are overcome. This allows for a more realistic reflection of the comprehensive impact of complex meteorological environments on wind turbine loads, thereby enabling a more comprehensive and accurate assessment of load asymmetry risks.

[0059] By establishing a multi-parameter coupled model of meteorological parameters and dynamic response of key components, it is possible to quantitatively analyze the collaborative stress state of key components such as blades, nacelles, and foundations under various meteorological conditions. This breaks through the limitations of the original two-dimensional risk assessment that only focuses on directional mismatch and wave frequency band, and provides more refined data support for the formulation of control strategies.

[0060] Based on a multi-dimensional risk assessment index system, with the goal of optimizing power generation efficiency, structural safety and component fatigue life globally, the yaw speed and pitch rate are optimized in a coordinated manner. This enables the wind turbine to not only alleviate instantaneous asymmetrical loads when dealing with extreme wind and wave mismatch conditions, but also balance power generation performance and structural damage during long-term operation, achieving overall optimization of safety and efficiency.

[0061] By collecting operational status feedback data in real time and dynamically correcting model parameter weights, the control system has the ability to continuously learn and adapt, improving control accuracy and robustness in complex scenarios such as sudden changes in meteorological parameters or multiple disasters, and ensuring stable and reliable operation of wind turbines in harsh marine environments. Attached Figure Description

[0062] Figure 1 This is a flowchart illustrating the wind turbine control method based on meteorological information according to an embodiment of the present invention.

[0063] Figure 2 This is a schematic diagram of the structure of a wind turbine control system based on meteorological information according to an embodiment of the present invention. Detailed Implementation

[0064] The following is a detailed description of the embodiments, in conjunction with the accompanying drawings.

[0065] Reference Figure 1 A wind turbine control method based on meteorological information includes the following steps:

[0066] Step 1: Real-time synchronous acquisition of multi-dimensional data, including basic load correlation data and comprehensive meteorological parameters; wherein, the basic load correlation data includes wind direction, wave direction, vibration signal of support structure and movement trajectory of tower top; the comprehensive meteorological parameters include wind speed turbulence intensity, atmospheric pressure fluctuation, precipitation intensity and temperature gradient.

[0067] In some embodiments, the specific process of real-time synchronous acquisition of multi-dimensional data is as follows:

[0068] By deploying sensor arrays in the wind turbine and its marine environment, basic load correlation data and comprehensive meteorological parameters are collected simultaneously;

[0069] The acquisition of the foundation load correlation data includes: obtaining wind direction and wave direction through wind direction sensor and wave direction sensor respectively; acquiring vibration signal of the support structure through acceleration sensor installed on the support structure; and recording the movement trajectory of the tower top through tower top positioning system;

[0070] The collection of comprehensive meteorological parameters includes: obtaining wind speed and calculating its turbulence intensity through an anemometer; monitoring atmospheric pressure fluctuations through a barometer; obtaining precipitation intensity through a rain gauge; and measuring temperature at different altitudes and calculating temperature gradients through a temperature sensor array.

[0071] The synchronously collected basic load correlation data and comprehensive meteorological parameters are integrated into multi-dimensional data with time-series labels.

[0072] Step 2: Standardize the multi-dimensional data to obtain a standardized dataset, and perform spectral analysis on the vibration signal of the standardized support structure to identify the dominant frequency components of wave excitation. Determine the concentrated frequency band of wave spectrum energy based on the standardized tower top motion trajectory.

[0073] In some embodiments, the time-series labels of multi-dimensional data can be used to precisely align the basic load-related data with comprehensive meteorological parameters on the time axis, ensuring that the wind direction, wave direction, support structure vibration signal, tower top movement trajectory and wind speed turbulence intensity, atmospheric pressure fluctuation, precipitation intensity and temperature gradient data at the same time stamp correspond one-to-one, forming a time-series aligned dataset.

[0074] The 3σ criterion was used to remove outliers from the time-series aligned dataset. Outlier data points exceeding the mean ± 3 standard deviations in each data dimension were filtered and removed to obtain the preprocessed dataset.

[0075] Based on data characteristics, the preprocessed dataset is divided into scalar data and time-series data. The scalar data includes wind direction angle values, wave direction angle values, wind speed turbulence intensity values, atmospheric pressure fluctuation values, precipitation intensity values, and temperature gradient values. The time-series data includes the time-domain signal of the support structure vibration and the time-domain sequence of the tower top motion trajectory.

[0076] Standardization operations are performed on scalar data and time series data respectively: For scalar data, the min-max normalization algorithm is used to map each data value to the [0,1] interval to eliminate the dimensional differences of different physical quantities; For time series data, the z-score standardization algorithm is used to calculate the deviation of each time domain data point from the corresponding sequence mean, and divide it by the sequence standard deviation to obtain the dimensionless standardized time domain signal.

[0077] The scalar data and time-series data that have undergone standardized operations are re-integrated according to time-series labels to form a standardized dataset that includes standardized data related to basic loads and standardized data of comprehensive meteorological parameters. The standardized dataset can be directly used for spectral analysis of vibration signals of supporting structures and determination of the concentrated frequency band of wave spectrum energy.

[0078] Step 3: Based on the standardized dataset, establish a multi-parameter coupling model of meteorological multi-parameters and dynamic response of key components of wind turbine. The multi-parameter coupling model takes the standardized basic load correlation data and comprehensive meteorological parameters as input and outputs the quantitative parameters of the coordinated stress state of each key component.

[0079] In some embodiments, the specific process of establishing a multi-parameter coupling model of meteorological multi-parameters and the dynamic response of key components of wind turbines based on a standardized dataset is as follows:

[0080] The standardized dataset is used as the model input, which includes a standardized subset of basic load-related data and a subset of comprehensive meteorological parameters. The model input undergoes nonlinear transformation and feature fusion through at least one hidden layer, which is used to extract high-dimensional coupling features between meteorological parameters and key component responses. Based on the fused features processed by the hidden layer, the output layer calculates and outputs the quantified parameters of the collaborative stress state of each key component, including blades, nacelles, and foundations. The quantified parameters of the collaborative stress state are vectors characterizing the dynamic mechanical responses of each component. The multi-parameter coupled model is trained using a historical operational data sample set. The weight parameters of each connection layer in the model are iteratively optimized by forward propagation calculation output and backpropagation error correction until the error between the quantified parameters of the collaborative stress state output by the model and the true values ​​of the samples is lower than a preset threshold.

[0081] We can set the standardized dataset as follows: ,in, This is a subset of the basic load-related data; This is a subset of comprehensive meteorological parameters. The key component set is... These correspond to the blades, nacelle, and foundation, respectively. The multi-parameter coupled model... It is represented as follows:

[0082]

[0083] in, Quantize the parameter vector for the output cooperative force state; For components The corresponding quantization parameter vector, for example, It is the quantization parameter vector of the blade; This is a functional representation of a multi-parameter coupled model; and These are the weight matrix and bias vector of the model, respectively.

[0084] Nonlinear mapping can be achieved through activation functions Achieve, the first Hidden layer output As shown below:

[0085]

[0086] in, Indicates the layer index of the neural network; is the activation function used to introduce nonlinear transformations into the hidden layers of the model, enabling the model to learn and fit complex nonlinear relationships (such as the relationship between meteorological parameters and structural responses). ; For the first Layer weight matrix; For the first The bias vector of the layer.

[0087] By minimizing the loss function To optimize parameters, These are the model's predicted values. The optimization objective is as follows, taking the true values ​​of the samples as the reference:

[0088]

[0089] in, The number of samples; Let L2 be the L2 norm of the vector. Iterative updates of the weights can be implemented using the gradient descent algorithm, as shown below:

[0090]

[0091] in, The learning rate controls the model parameters in each iteration. and Update step size.

[0092] Step 4: Based on the quantitative parameters of the coordinated stress state, combined with the dominant frequency components of wave excitation, the concentrated frequency band of wave spectrum energy, and the instantaneous difference between wind direction and wave direction, a multi-dimensional risk assessment index system is constructed to calculate the load asymmetry risk level of each key component.

[0093] In some embodiments, the specific process of constructing a multi-dimensional risk assessment index system and calculating the load asymmetry risk level of each key component is as follows:

[0094] Based on the aforementioned quantified parameters of the coordinated stress state, a load asymmetry evaluation factor is defined to characterize the degree of dynamic load imbalance in key components. The mean and standard deviation of the quantified parameters of the coordinated stress state, output by the multi-parameter coupled model, can be calculated as follows:

[0095]

[0096] in, Indicates the first The arithmetic mean of the quantized parameter vectors of the collaborative stress state corresponding to each key component; Indicates the first The standard deviation of the quantification parameter vector of the collaborative stress state corresponding to each key component reflects the degree of fluctuation or dispersion of the dynamic load of the component around the average value. The larger the value, the more drastic the load change. Indicates the first The first parameter in the quantization parameter vector of the collaborative stress state of the key components The element (i.e., the element) (one quantization parameter) The dimension of the parameter vector for the cooperative stress state quantization, i.e., the number of quantization parameters contained in the vector;

[0097] The load asymmetry evaluation factor is defined as follows:

[0098]

[0099] in, For the first Evaluation factors for load asymmetry of key components Dimensionless, the larger the value, the more components it represents. The higher the degree of dynamic load imbalance; Represents an extremely small positive number (e.g., 1e-6) in calculations It was introduced at the time to prevent the denominator from being zero.

[0100] Based on the dominant frequency components of wave excitation, the concentrated frequency band of wave spectrum energy, and the instantaneous difference between wind direction and wave direction, wave excitation frequency evaluation factors, wave energy concentration evaluation factors, and wind-wave direction mismatch evaluation factors are defined respectively. It can be set that... To identify the dominant wave excitation frequency from spectral analysis, If the reference natural frequency of the wind turbine support structure or key components (which can be obtained through modal analysis) is given, the wave excitation frequency evaluation factor is defined as follows:

[0101]

[0102] in, As a wave excitation frequency evaluation factor, ; This is a sensitivity parameter used to adjust the evaluation factors. For frequency difference ( The sensitivity of ( ) is determined based on historical operational data. When near hour, A value close to 1 indicates a high risk of resonance; conversely, a value close to 0 indicates a low risk of resonance.

[0103] Let the wave spectrum energy concentration band be defined. The bandwidth is Energy concentration coefficient (This can be calculated by the ratio of peak energy to total energy in the wave spectrum), and the wave energy concentration assessment factor is defined as follows:

[0104]

[0105] in, As a factor for assessing wave energy concentration, Dimensionless; the larger the value, the more concentrated the wave energy and the higher the risk. For weight parameters (satisfying) ); For a very small positive number, in Add a denominator to the calculation formula to prevent division by zero errors.

[0106] set up The wind direction angle (in radians). The instantaneous difference represents the wave direction angle (in radians). The wind and wave direction mismatch assessment factor is defined as follows:

[0107]

[0108] in, For the assessment factor of wind and wave direction mismatch. The larger the value, the more severe the mismatch between wind and wave directions.

[0109] The load asymmetry assessment factor, wave excitation frequency assessment factor, wave energy concentration assessment factor, and wind-wave direction mismatch assessment factor are used as the four core assessment factors in the multi-dimensional risk assessment index system. Weight coefficients are assigned to these four core assessment factors, and these weight coefficients are determined based on the correlation analysis between historical meteorological data and component failure data. Based on the assigned weight coefficients, a weighted fusion calculation is performed on the four core assessment factors to obtain the comprehensive risk assessment value for each key component. This can be applied to each component... The four core evaluation factors are assigned weight coefficients. These weights are determined through correlation analysis (such as multiple regression or machine learning) between historical meteorological data and component failure data, and satisfy normalization conditions, as shown in the following formula:

[0110]

[0111] in, Indicates the first The first key component corresponding to the The weighting coefficients of each core evaluation factor; Corresponding to , , and These four evaluation factors, and their weighting coefficients, were determined through correlation analysis of historical meteorological data and component failure data, and met the conditions of normalization and non-negativity.

[0112] Then the component The comprehensive risk assessment value is shown in the following formula:

[0113]

[0114] in, Indicates the first Comprehensive risk assessment value of each key component.

[0115] Based on the preset risk level threshold range, the comprehensive risk assessment value is mapped to the corresponding load asymmetry risk level.

[0116] In some embodiments, after the multi-dimensional risk assessment index system calculates the load asymmetry risk level of each key component, the method further includes the following steps:

[0117] When the load asymmetry risk level of any critical component exceeds the preset severe risk threshold, a simulation sequence of the evolution of meteorological parameters towards extreme working conditions in future time periods is generated based on the standardized dataset and the multi-parameter coupling model at the current moment.

[0118] The simulation sequence is input into the multi-parameter coupled model to calculate the predictive collaborative stress state quantification parameters of each key component in the future time period;

[0119] Based on the predictive collaborative stress state quantification parameters, with loss prevention and power generation maintenance as the core optimization objectives, the preventive collaborative adjustment parameter sequence for the future period is generated in advance through iterative optimization.

[0120] The preventive coordinated adjustment parameter sequence is stored in the control queue. When the actual meteorological conditions change to be close to the simulation sequence, the preventive coordinated adjustment parameters in the control queue are called in sequence and sent to the wind turbine actuator to perform advanced control.

[0121] Step 5: Based on the load asymmetry risk level, and with the goal of achieving the global optimal power generation efficiency, structural safety, and component fatigue life, optimize and generate coordinated adjustment parameters for yaw speed and pitch rate.

[0122] In some embodiments, the specific process of optimizing the coordinated adjustment parameters of yaw speed and pitch rate based on the load asymmetry risk level, with the goal of achieving global optimization of power generation efficiency, structural safety, and component fatigue life, is as follows:

[0123] Based on the load asymmetry risk level, the basic yaw speed adjustment coefficient and the basic pitch rate adjustment coefficient are queried and determined from the preset risk level-adjustment strategy mapping table.

[0124] The basic yaw speed adjustment coefficient and the basic pitch rate adjustment coefficient are fused with the quantified parameters of the cooperative stress state output by the multi-parameter coupling model and input to the preset multi-objective optimization controller; the multi-objective optimization controller has a built-in optimization function with the sub-objectives of maximizing power generation efficiency, minimizing structural safety risk, and minimizing component fatigue damage;

[0125] Based on the quantized parameters of the cooperative stress state, the multi-objective optimization controller calculates, under the current stress state, the contribution of adjusting the yaw speed to power generation efficiency, the suppression coefficient of structural safety risk, and the mitigation weight of adjusting the pitch rate for fatigue damage.

[0126] Based on the contribution, suppression coefficient and mitigation weight, the basic yaw speed adjustment coefficient and basic pitch rate adjustment coefficient are weighted and corrected to generate preliminary coordinated adjustment parameters.

[0127] Using the initial coordinated adjustment parameters as the initial solution, a multi-objective optimization algorithm is called to iteratively solve the optimization function. Under the condition of satisfying the dynamic load constraints of each key component, the final yaw speed command value and the final pitch rate command value that achieve the global optimal balance between power generation efficiency, structural safety and component fatigue life are obtained.

[0128] The final yaw speed command value and the final pitch rate command value are integrated into the coordinated adjustment parameters.

[0129] Step 6: Send the coordinated adjustment parameters to the wind turbine actuator for execution control, collect real-time feedback data on the wind turbine's operating status, and send it back to the multi-parameter coupled model to dynamically correct the model parameter weights.

[0130] In some embodiments, the specific process of real-time acquisition of wind turbine operating status feedback data and transmission back to the multi-parameter coupled model to dynamically correct the model parameter weights is as follows:

[0131] The operational status feedback data includes real-time stress data of key components, wind turbine speed data, yaw actuator response speed data, and pitch actuator response angle data.

[0132] The deviation between the quantified parameters of the coordinated stress state output by the multi-parameter coupled model and the real-time stress data of key components in the operation status feedback data is calculated. If the deviation exceeds the preset deviation threshold, the model parameter weight correction process is triggered.

[0133] Based on the magnitude and direction of the deviation, the gradient descent method is used to dynamically adjust the connection weights of the basic load associated data input layer, the comprehensive meteorological parameter input layer and the hidden layer in the multi-parameter coupled model. During the adjustment process, the weight parameters are iteratively updated with the goal of minimizing the deviation until the deviation is lower than the preset deviation threshold.

[0134] The adjusted weight parameters are saved to the model parameter library as the latest weight benchmark for subsequent multi-parameter coupled model calculations, and the time series information and corresponding meteorological conditions of this correction are recorded.

[0135] Based on the same inventive concept, and corresponding to any of the above embodiments, refer to... Figure 2 This invention provides a wind turbine control system based on meteorological information, used to implement the aforementioned wind turbine control method based on meteorological information, comprising:

[0136] The data synchronization acquisition module is used to collect multi-dimensional data in real time through a sensor array deployed on the wind turbine and its marine environment. The multi-dimensional data includes basic load correlation data and comprehensive meteorological parameters. The basic load correlation data includes wind direction, wave direction, vibration signal of the support structure, and movement trajectory of the tower top. The comprehensive meteorological parameters include wind speed turbulence intensity, atmospheric pressure fluctuation, precipitation intensity, and temperature gradient.

[0137] The data processing and analysis module is connected to the data synchronization acquisition module. It is used to standardize the received multi-dimensional data to obtain a standardized dataset. It also performs spectrum analysis based on the standardized support structure vibration signal to identify the dominant frequency component of wave excitation, and determines the concentrated frequency band of wave spectrum energy based on the standardized tower top motion trajectory.

[0138] The model building and calculation module is connected to the data processing and analysis module. It is used to build and run a multi-parameter coupled model of meteorological multi-parameters and dynamic response of key components of wind turbine based on the standardized dataset. The model takes standardized basic load correlation data and comprehensive meteorological parameters as input, and outputs the quantified parameters of the coordinated stress state of each key component after nonlinear transformation and feature fusion through at least one hidden layer.

[0139] The risk assessment module is connected to the data processing and analysis module and the model building and calculation module, respectively. It is used to construct a multi-dimensional risk assessment index system based on the quantitative parameters of the coordinated stress state, combined with the dominant frequency components of wave excitation, the concentrated frequency band of wave spectrum energy and the instantaneous difference between wind direction and wave direction, and to calculate the load asymmetry risk level of each key component.

[0140] The collaborative control parameter optimization module is connected to the risk assessment module and the model construction and calculation module; it is used to optimize and generate collaborative adjustment parameters for yaw speed and pitch rate by calling a multi-objective optimization algorithm based on the load asymmetry risk level and with the global optimization of power generation efficiency, structural safety and component fatigue life as the objectives.

[0141] The execution control and feedback module is connected to the collaborative control parameter optimization module; it is used to send the collaborative adjustment parameters to the yaw and pitch actuators of the wind turbine to perform control; and to collect the operating status feedback data of the wind turbine in real time and send it back to the model building and calculation module to dynamically correct the parameter weights of the multi-parameter coupled model.

[0142] Based on the same inventive concept, corresponding to any of the above embodiments, the present invention provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the wind turbine control method based on meteorological information of the embodiment.

[0143] Alternatively, the aforementioned electronic device may be a server.

[0144] In addition, this embodiment also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the wind turbine control method based on meteorological information of the embodiment.

[0145] It is understood that the processor in the embodiments of the present invention may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0146] The method steps in the embodiments of the present invention can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0147] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a storage medium or transmitted through a storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive (SSD)).

Claims

1. A wind turbine control method based on meteorological information, characterized in that, Includes the following steps: Real-time synchronous acquisition of multi-dimensional data, including basic load correlation data and comprehensive meteorological parameters; wherein, the basic load correlation data includes wind direction, wave direction, vibration signal of support structure and movement trajectory of tower top; the comprehensive meteorological parameters include wind speed turbulence intensity, atmospheric pressure fluctuation, precipitation intensity and temperature gradient; Multidimensional data is standardized to obtain a standardized dataset. Spectral analysis is then performed on the standardized vibration signal of the support structure to identify the dominant frequency components of wave excitation. The concentrated frequency band of wave spectrum energy is determined based on the standardized tower top motion trajectory. Based on a standardized dataset, a multi-parameter coupling model of meteorological multi-parameters and dynamic response of key components of wind turbine is established. The multi-parameter coupling model takes standardized basic load correlation data and comprehensive meteorological parameters as input and outputs the quantitative parameters of the coordinated stress state of each key component. Based on the quantitative parameters of the coordinated stress state, combined with the dominant frequency components of wave excitation, the concentrated frequency band of wave spectrum energy, and the instantaneous difference between wind direction and wave direction, a multi-dimensional risk assessment index system is constructed to calculate the load asymmetry risk level of each key component. Based on the load asymmetry risk level, and with the goal of achieving the global optimum of power generation efficiency, structural safety, and component fatigue life, the coordinated adjustment parameters for yaw speed and pitch rate are optimized and generated. The coordinated adjustment parameters are sent to the wind turbine actuator for execution control, and the wind turbine operating status feedback data is collected in real time and sent back to the multi-parameter coupled model to dynamically correct the model parameter weights.

2. The wind turbine control method based on meteorological information as described in claim 1, characterized in that... The specific process of real-time synchronous acquisition of multi-dimensional data is as follows: By deploying sensor arrays in the wind turbine and its marine environment, basic load correlation data and comprehensive meteorological parameters are collected simultaneously; The acquisition of the foundation load correlation data includes: obtaining wind direction and wave direction through wind direction sensor and wave direction sensor respectively; acquiring vibration signal of the support structure through acceleration sensor installed on the support structure; and recording the movement trajectory of the tower top through tower top positioning system; The collection of comprehensive meteorological parameters includes: obtaining wind speed and calculating its turbulence intensity through an anemometer; monitoring atmospheric pressure fluctuations through a barometer; obtaining precipitation intensity through a rain gauge; and measuring temperature at different altitudes and calculating temperature gradients through a temperature sensor array. The synchronously collected basic load correlation data and comprehensive meteorological parameters are integrated into multi-dimensional data with time-series labels.

3. The wind turbine control method based on meteorological information as described in claim 1, characterized in that, The specific process of establishing a multi-parameter coupled model of meteorological multi-parameters and the dynamic response of key components of wind turbines based on a standardized dataset is as follows: The standardized dataset is used as the model input, which includes a standardized subset of basic load correlation data and a subset of comprehensive meteorological parameters. The model input is subjected to nonlinear transformation and feature fusion through at least one hidden layer, which is used to extract high-dimensional coupling features between meteorological parameters and key component responses; Based on the fusion features processed by the hidden layer, the output layer calculates and outputs the quantification parameters of the collaborative stress state of each key component, including blades, nacelle and foundation. The quantification parameters of the collaborative stress state are vectors characterizing the dynamic mechanical response of each component. The multi-parameter coupled model is trained using a historical operational data sample set. The weight parameters of each connection layer in the model are iteratively optimized by calculating the output of forward propagation and correcting the error of backward propagation, until the error between the model output of the quantified parameters of the cooperative force state and the true value of the sample is lower than a preset threshold.

4. The wind turbine control method based on meteorological information as described in claim 1, characterized in that, The specific process for constructing a multi-dimensional risk assessment index system and calculating the load asymmetry risk level of each key component is as follows: Based on the aforementioned quantification parameters of the coordinated stress state, a load asymmetry evaluation factor is defined to characterize the degree of dynamic load imbalance of key components. Based on the dominant frequency component of wave excitation, the concentrated frequency band of wave spectrum energy, and the instantaneous difference between wind direction and wave direction, wave excitation frequency evaluation factor, wave energy concentration evaluation factor, and wind-wave direction mismatch evaluation factor are defined respectively. The load asymmetry assessment factor, the wave excitation frequency assessment factor, the wave energy concentration assessment factor, and the wind and wave direction mismatch assessment factor are used as the four core assessment factors of the multi-dimensional risk assessment index system. Weight coefficients are assigned to the four core assessment factors, and the weight coefficients are determined based on the correlation analysis between historical meteorological data and component failure data. Based on the assigned weight coefficients, the four core assessment factors are weighted and fused to obtain the comprehensive risk assessment value of each key component; Based on the preset risk level threshold range, the comprehensive risk assessment value is mapped to the corresponding load asymmetry risk level.

5. The wind turbine control method based on meteorological information as described in claim 4, characterized in that, After calculating the load asymmetry risk level of each key component using the multi-dimensional risk assessment index system, the following steps are also included: When the load asymmetry risk level of any critical component exceeds the preset severe risk threshold, a simulation sequence of the evolution of meteorological parameters towards extreme working conditions in future time periods is generated based on the standardized dataset and the multi-parameter coupling model at the current moment. The simulation sequence is input into the multi-parameter coupled model to calculate the predictive collaborative stress state quantification parameters of each key component in the future time period; Based on the predictive collaborative stress state quantification parameters, with loss prevention and power generation maintenance as the core optimization objectives, the preventive collaborative adjustment parameter sequence for the future period is generated in advance through iterative optimization. The preventive coordinated adjustment parameter sequence is stored in the control queue. When the actual meteorological conditions change to be close to the simulation sequence, the preventive coordinated adjustment parameters in the control queue are called in sequence and sent to the wind turbine actuator to perform advanced control.

6. The wind turbine control method based on meteorological information as described in claim 1, characterized in that, The specific process of optimizing the coordinated adjustment parameters of yaw speed and pitch rate based on the load asymmetry risk level, with the goal of achieving global optimization of power generation efficiency, structural safety, and component fatigue life, is as follows: Based on the load asymmetry risk level, the basic yaw speed adjustment coefficient and the basic pitch rate adjustment coefficient are queried and determined from the preset risk level-adjustment strategy mapping table. The basic yaw speed adjustment coefficient and the basic pitch rate adjustment coefficient are fused with the quantified parameters of the cooperative stress state output by the multi-parameter coupling model and input to the preset multi-objective optimization controller; the multi-objective optimization controller has a built-in optimization function with the sub-objectives of maximizing power generation efficiency, minimizing structural safety risk, and minimizing component fatigue damage; Based on the quantized parameters of the cooperative stress state, the multi-objective optimization controller calculates, under the current stress state, the contribution of adjusting the yaw speed to power generation efficiency, the suppression coefficient of structural safety risk, and the mitigation weight of adjusting the pitch rate for fatigue damage. Based on the contribution, suppression coefficient and mitigation weight, the basic yaw speed adjustment coefficient and basic pitch rate adjustment coefficient are weighted and corrected to generate preliminary coordinated adjustment parameters. Using the initial coordinated adjustment parameters as the initial solution, a multi-objective optimization algorithm is called to iteratively solve the optimization function. Under the condition of satisfying the dynamic load constraints of each key component, the final yaw speed command value and the final pitch rate command value that achieve the global optimal balance between power generation efficiency, structural safety and component fatigue life are obtained. The final yaw speed command value and the final pitch rate command value are integrated into the coordinated adjustment parameters.

7. The wind turbine control method based on meteorological information as described in claim 1, characterized in that, The specific process of collecting real-time wind turbine operating status feedback data and transmitting it back to the multi-parameter coupled model to dynamically correct the model parameter weights is as follows: The operational status feedback data includes real-time stress data of key components, wind turbine speed data, yaw actuator response speed data, and pitch actuator response angle data. The deviation between the quantified parameters of the coordinated stress state output by the multi-parameter coupled model and the real-time stress data of key components in the operation status feedback data is calculated. If the deviation exceeds the preset deviation threshold, the model parameter weight correction process is triggered. Based on the magnitude and direction of the deviation, the gradient descent method is used to dynamically adjust the connection weights of the basic load associated data input layer, the comprehensive meteorological parameter input layer and the hidden layer in the multi-parameter coupled model. During the adjustment process, the weight parameters are iteratively updated with the goal of minimizing the deviation until the deviation is lower than the preset deviation threshold. The adjusted weight parameters are saved to the model parameter library as the latest weight benchmark for subsequent multi-parameter coupled model calculations, and the time series information and corresponding meteorological conditions of this correction are recorded.

8. A wind turbine control system based on meteorological information, used to implement the wind turbine control method based on meteorological information as described in any one of claims 1-7, characterized in that, include: The data synchronization acquisition module is used to collect multi-dimensional data in real time through a sensor array deployed on the wind turbine and its marine environment. The multi-dimensional data includes basic load correlation data and comprehensive meteorological parameters; wherein, the basic load correlation data includes wind direction, wave direction, vibration signal of supporting structure and movement trajectory of tower top; the comprehensive meteorological parameters include wind speed turbulence intensity, atmospheric pressure fluctuation, precipitation intensity and temperature gradient; The data processing and analysis module is connected to the data synchronization acquisition module. It is used to standardize the received multi-dimensional data to obtain a standardized dataset. It also performs spectrum analysis based on the standardized support structure vibration signal to identify the dominant frequency component of wave excitation, and determines the concentrated frequency band of wave spectrum energy based on the standardized tower top motion trajectory. The model building and calculation module is connected to the data processing and analysis module. It is used to build and run a multi-parameter coupled model of meteorological multi-parameters and dynamic response of key components of wind turbine based on the standardized dataset. The model takes standardized basic load correlation data and comprehensive meteorological parameters as input, and outputs the quantified parameters of the coordinated stress state of each key component after nonlinear transformation and feature fusion through at least one hidden layer. The risk assessment module is connected to the data processing and analysis module and the model building and calculation module, respectively. It is used to construct a multi-dimensional risk assessment index system based on the quantitative parameters of the coordinated stress state, combined with the dominant frequency components of wave excitation, the concentrated frequency band of wave spectrum energy and the instantaneous difference between wind direction and wave direction, and to calculate the load asymmetry risk level of each key component. The collaborative control parameter optimization module is connected to the risk assessment module and the model construction and calculation module; it is used to optimize and generate collaborative adjustment parameters for yaw speed and pitch rate by calling a multi-objective optimization algorithm based on the load asymmetry risk level and with the global optimization of power generation efficiency, structural safety and component fatigue life as the objectives. The execution control and feedback module is connected to the collaborative control parameter optimization module; it is used to send the collaborative adjustment parameters to the yaw and pitch actuators of the wind turbine to perform control; and to collect the operating status feedback data of the wind turbine in real time and send it back to the model building and calculation module to dynamically correct the parameter weights of the multi-parameter coupled model.

9. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store a computer program, and the processor running the computer program to cause the electronic device to perform the wind turbine control method based on meteorological information as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the wind turbine control method based on meteorological information as described in any one of claims 1-7.

Images