Intelligent variable pitch control method, system and equipment for wind generating set

Abstract

The invention discloses an intelligent variable pitch control method, system and equipment for a wind generating set, and relates to the technical field of wind generating set control, and the method comprises the steps: synchronously collecting multi-source data in real time through a multi-source heterogeneous sensor array, constructing a dynamic load spectrum matrix, and carrying out nonlinear optimization; an optimal pitch angle sequence is generated to drive a variable pitch execution mechanism, meanwhile, the blade stress change rate is monitored in real time, and an aerodynamic load-structure response coupling factor is calculated; when the factor exceeds a safety threshold, the pitch angle command is reconstructed and the pitch angle is forcibly adjusted to a preset safety angle. The technical problems that the power generation efficiency and the structure safety are difficult to consider under the complex turbulence condition and a safety protection mechanism under the over-limit working condition is lacked in the variable pitch control of the existing wind generating set are solved, and the aim that the power generation efficiency and the structure safety are effectively improved through nonlinear optimization control and introduction of an over-limit triggered standby control path is achieved. And the technical effect of dynamic unification of power generation efficiency and structural safety in a complex turbulence environment is achieved.

Description

Technical Field

[0001] This invention relates to the field of wind turbine control technology, specifically to intelligent pitch control methods, systems, and equipment for wind turbine generator sets. Background Technology

[0002] In existing wind turbine operation control, pitch control is a key means of regulating output power and reducing structural load. However, existing control methods mostly rely on a single wind speed or power signal for adjustment, and the control logic is relatively simple. While it can meet basic requirements under steady-state wind conditions, it has significant shortcomings in actual complex turbulent environments. For example, on the one hand, the rapid transient changes in turbulent wind speed lead to uneven stress on the blades, easily causing local stress concentration and accelerating fatigue damage to the blades and bearings. On the other hand, existing linear model-based control methods cannot accurately reflect the nonlinear coupling relationship between wind turbine aerodynamic disturbances and structural response, and control commands are at risk of lag or over-adjustment. In addition, existing methods lack safety redundancy mechanisms under extreme wind conditions, making the unit vulnerable to operational instability, stall, or even structural failure when subjected to sudden load impacts. Summary of the Invention

[0003] This application provides a method, system, and equipment for intelligent pitch control of wind turbine generator sets, which solves the technical problems that existing pitch control of wind turbine generator sets is difficult to balance power generation efficiency and structural safety under complex turbulent conditions, and lacks a safety protection mechanism under over-limit operating conditions.

[0004] The first aspect of this application provides an intelligent pitch control method for wind turbine generator sets. The method includes: synchronously acquiring turbulent wind speed, blade root strain energy density, generator instantaneous power, and pitch bearing temperature in real time using a multi-source heterogeneous sensor array to construct a dynamic load spectrum matrix; inputting the dynamic load spectrum matrix into a coupling control unit for nonlinear optimization to generate an optimal pitch angle sequence; driving the pitch actuator according to the optimal pitch angle sequence while simultaneously monitoring the blade stress change rate in real time and calculating the aerodynamic load-structure response coupling factor; when the aerodynamic load-structure response coupling factor exceeds a safety threshold, triggering a backup control path to reconstruct the pitch angle command, and forcibly adjusting the pitch angle to a preset safety angle according to the reconstructed pitch angle sequence.

[0005] A second aspect of this application provides an intelligent pitch control system for wind turbine generator sets. The system includes: a multi-source data acquisition module for real-time synchronous acquisition of turbulent wind speed, blade root strain energy density, generator instantaneous power, and pitch bearing temperature via a multi-source heterogeneous sensor array, constructing a dynamic load spectrum matrix; a nonlinear optimization module for inputting the dynamic load spectrum matrix into a coupling control unit for nonlinear optimization, generating an optimal pitch angle sequence; a dynamic monitoring factor calculation module for driving the pitch actuator according to the optimal pitch angle sequence, while simultaneously monitoring the blade stress change rate in real time, and calculating the aerodynamic load-structure response coupling factor; and a pitch angle reconstruction module for triggering a backup control path to reconstruct the pitch angle command when the aerodynamic load-structure response coupling factor exceeds a safety threshold, and forcibly adjusting the pitch angle to a preset safe angle according to the reconstructed pitch angle sequence.

[0006] A third aspect of this application provides an electronic device comprising: a processor coupled to a memory for storing a program that, when executed by the processor, causes the system to perform the method described in any of the first aspects.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The intelligent pitch control method, system, and equipment for wind turbine generator sets provided in this application relate to the field of wind turbine control technology. It constructs a dynamic load spectrum matrix by fusing multi-source heterogeneous sensor data and generates an optimal pitch angle sequence through nonlinear optimization combining feedforward prediction and feedback compensation. Simultaneously, it introduces an aerodynamic load-structural response coupling factor as a safety criterion. When the limit is exceeded, a reconfigured pitch angle command is triggered and forced to adjust to a safe angle, achieving a dynamic balance between power generation efficiency and structural safety. This solves the technical problem that existing wind turbine generator pitch control struggles to balance power generation efficiency and structural safety under complex turbulent conditions and lacks a safety protection mechanism under over-limit conditions. It achieves the technical effect of dynamically unifying power generation efficiency and structural safety in complex turbulent environments through multi-source data fusion and nonlinear optimization control, and by introducing a backup control path for over-limit triggering. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic flowchart of an intelligent pitch control method for wind turbine generator sets provided in an embodiment of this application.

[0011] Figure 2 This is a schematic diagram of the intelligent pitch control system for wind turbine generators provided in an embodiment of this application;

[0012] Figure 3 This application provides a schematic diagram of the structure of an electronic device.

[0013] Figure reference numerals: 11 Multi-source data acquisition module, 12 Nonlinear optimization module, 13 Dynamic monitoring factor calculation module, 14 Pitch angle reconstruction module, 300 Electronic device, 301 Memory, 302 Processor, 303 Communication interface, 304 Bus architecture. Detailed Implementation

[0014] This application provides a method, system, and equipment for intelligent pitch control of wind turbine generator sets, which solves the technical problems that existing pitch control of wind turbine generator sets is difficult to balance power generation efficiency and structural safety under complex turbulent conditions, and lacks a safety protection mechanism under over-limit operating conditions.

[0015] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0016] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0017] Example 1, as Figure 1 As shown, this application provides an intelligent pitch control method for wind turbine generator sets, the method comprising:

[0018] P10: A dynamic load spectrum matrix is ​​constructed by synchronously collecting turbulent wind speed, blade root strain energy density, generator instantaneous power, and pitch bearing temperature in real time through a multi-source heterogeneous sensor array.

[0019] Furthermore, step P10 in this embodiment of the application also includes:

[0020] P11: Perform frequency domain feature decomposition on the collected turbulent wind speed to obtain the feature components characterizing the turbulence intensity; P12: Sample the strain energy density at the blade root in equal circumferential angles to generate a strain spatial distribution vector; P13: Perform time-series correlation mapping between the instantaneous power of the generator and the pitch bearing temperature to form a power-temperature correlation matrix; P14: Fuse the aforementioned feature components, strain spatial distribution vector, and power-temperature correlation matrix to generate a dynamic load spectrum matrix.

[0021] It should be understood that the operating conditions of the wind turbine generator are first collected in real time and synchronously using a multi-source heterogeneous sensor array. This multi-source heterogeneous sensor array includes, but is not limited to: wind speed sensors for detecting turbulent wind speed, such as laser Doppler anemometers or ultrasonic anemometers; strain gauges or fiber Bragg grating sensors for measuring strain energy density at the blade root; power sensors for acquiring instantaneous generator power; and thermocouples or infrared temperature sensors for monitoring pitch bearing temperature. All of these sensors need to be synchronized with a unified clock module to ensure that all types of sensors complete signal acquisition under the same time reference, thereby achieving consistency in the time series at the data level.

[0022] Next, frequency domain feature decomposition is performed on the collected turbulent wind speed. Turbulent wind speed is an important input parameter during the operation of wind turbine generators, and its complexity and randomness have a significant impact on the load and performance of the generator. Through frequency domain feature decomposition, the complex turbulent wind speed signal can be decomposed into several feature components. For example, signal processing algorithms such as Fast Fourier Transform, Short-Time Fourier Transform, or Wavelet Decomposition are used to decompose the wind speed time-domain signal into multiple frequency components, and the energy distribution corresponding to each frequency component is calculated. By comparing the energy proportion and attenuation law of different frequency bands, feature components that can reflect the turbulence intensity are extracted. Turbulence intensity is an important parameter for measuring the degree of wind field disturbance, and its feature components can quantitatively characterize the distribution of wind field energy in the high-frequency and low-frequency regions, thereby providing more accurate aerodynamic input indicators for subsequent control.

[0023] Next, the strain energy density at the blade root is sampled in circumferentially equal-angled zones to generate a strain spatial distribution vector. The strain energy density at the blade root is a crucial indicator of the blade's stress state. Since the load distribution on the blade during rotation is non-uniform, circumferentially equal-angled sampling allows for more detailed capture of the spatial distribution characteristics of the strain energy density at the blade root. For example, the 360° circumference of the blade root is divided into several equal segments, such as 12 or 24 angular regions, and strain energy density signals are acquired in real time within each region. Strain energy density refers to the elastic energy stored in a unit volume of material, and its magnitude is closely related to the stress level experienced by the material at that location. By vectorizing the sampling results of each zone, a strain spatial distribution vector is formed. This vector can intuitively characterize the stress differences at the blade root in different angular directions. Especially under non-uniform wind fields and complex turbulent conditions, it can reveal the spatial distribution characteristics of the blade's structural response, providing crucial information for structural load assessment and fatigue analysis.

[0024] Then, a time-series correlation mapping is performed between the generator's instantaneous power and the pitch bearing temperature to form a power-temperature correlation matrix. The generator's instantaneous power is a direct indicator of the wind turbine's power generation efficiency, while the pitch bearing temperature is a crucial parameter reflecting the operating status of the pitch system. This time-series correlation mapping establishes the intrinsic relationship between these two parameters. Specifically, the two types of data are first timestamped to ensure a one-to-one correspondence between data points. Subsequently, cross-correlation analysis, time-series regression models, or sliding window correlation calculation methods are used to reveal the correlation between power output fluctuations and bearing temperature changes, and a power-temperature correlation matrix is ​​constructed. This matrix quantifies the impact of power output changes on bearing thermal characteristics. For example, under conditions of frequent generator load fluctuations, the bearing temperature may exhibit a rapid upward trend, reflecting increased mechanical friction and wear. This matrix not only provides a reference for operational efficiency analysis but also serves as a basis for potential fault prediction and lifespan assessment.

[0025] Finally, the aforementioned feature components, strain spatial distribution vector, and power-temperature correlation matrix are fused to generate a dynamic load spectrum matrix. Specific fusion methods can include direct feature stitching, or principal component analysis, singular value decomposition, or deep learning feature extraction methods to reduce the dimensionality and weight the multi-dimensional features, thereby obtaining a comprehensive data representation in a unified coordinate system. By organically integrating data from different sources and dimensions, the dynamic load spectrum matrix can comprehensively and accurately reflect the various load conditions experienced by the wind turbine generator during operation. This matrix not only includes aerodynamic loads caused by wind speed changes but also covers information on blade structural stress distribution and the operating status of the pitch system, providing a rich data foundation and comprehensive load feature description for subsequent nonlinear optimization and pitch angle control.

[0026] P20: Input the dynamic load spectrum matrix into the coupling control unit for nonlinear optimization to generate the optimal pitch angle sequence.

[0027] Furthermore, step P20 in this embodiment of the application also includes:

[0028] P21: The coupling control unit includes a feedforward channel and a feedback channel; P22: The feedforward channel performs time-series prediction of the turbulent wind speed characteristics in the dynamic load spectrum matrix to generate a feedforward pitch angle sequence; P23: The feedback channel dynamically generates a compensation amount based on the generator power deviation in the dynamic load spectrum matrix; P24: The feedforward pitch angle sequence and the compensation amount are fused to output the optimal pitch angle sequence.

[0029] Optionally, the dynamic load spectrum matrix can be input into the coupled control unit for nonlinear optimization to nonlinearly optimize the pitch control of the wind turbine generator, thereby generating the optimal pitch angle sequence. Through nonlinear optimization calculations, the control objectives of maximizing power generation efficiency and minimizing structural loads can be simultaneously considered, enabling the wind turbine to operate safely, stably, and efficiently in complex turbulent wind fields.

[0030] The coupled control unit specifically includes a feedforward channel and a feedback channel. The feedforward channel primarily addresses wind speed disturbance characteristics in the dynamic load spectrum matrix, used to predict future operating condition changes in advance; the feedback channel compensates for deviations in real-time operating data. By combining the two, an organic coordination can be achieved between predictive control and real-time correction, making the control strategy both forward-looking and dynamically adaptable.

[0031] Specifically, the feedforward channel of the coupled control unit first performs time-series prediction of the turbulent wind speed characteristics in the dynamic load spectrum matrix. This time-series prediction of turbulent wind speed is based on historical and real-time monitoring data, using prediction algorithms (such as machine learning models or physical models) to predict wind speed trends over a future period. For example, the time-series components of the wind speed characteristics are first extracted, and then the prediction model is used to extrapolate the trend. For instance, autoregressive moving average (ARMA), long short-term memory neural network (LSTM), Kalman filtering, or spectral analysis-based prediction algorithms can be used to estimate wind speed fluctuations in the future. After obtaining the predicted wind speed characteristics, the feedforward channel calculates the feedforward pitch angle sequence based on the mapping relationship between wind speed and blade pitch angle, enabling the system to adjust the blade angle in advance before disturbances arrive, thereby reducing the impact of aerodynamic shocks on the unit.

[0032] Meanwhile, the feedback channel of the coupled control unit dynamically generates compensation based on the generator power deviation in the dynamic load spectrum matrix. Generator power deviation refers to the difference between the actual power output and the target power output; its magnitude reflects the degree of deviation between the output and the desired operating state. The feedback channel monitors this deviation and adjusts the pitch angle in real time to ensure that the generator's power output remains stable near the target value. For example, the feedback channel calculates this deviation in real time and uses proportional-integral-derivative (PID) control or fuzzy control algorithms to convert the power deviation into a corresponding compensation signal. This compensation can be directly used to correct the feedforward pitch angle sequence to avoid control deviations caused by errors in the prediction model or sudden changes in wind speed, thereby ensuring that the unit's operating power closely follows the target value.

[0033] Finally, the coupling control unit fuses the feedforward pitch angle sequence with the compensation amount generated by the feedback channel to output the optimal pitch angle sequence. The fusion process needs to comprehensively consider both prediction and real-time feedback information, balancing the requirements of advance adjustment and real-time correction. For example, a weighted superposition method can be used, adaptively adjusting the weights of feedforward and feedback according to different operating conditions; alternatively, a combination of nonlinear functions can be used to achieve smooth switching under dynamic conditions. Furthermore, an adaptive fusion algorithm based on machine learning can be introduced to train on long-term operating data, enabling the fusion strategy to automatically optimize according to different turbulence characteristics and power deviations. The final output optimal pitch angle sequence can be used as a control command input to the pitch actuator to achieve real-time adjustment of the blade angle, thereby improving the operating performance and structural safety of the wind turbine generator under unsteady wind conditions.

[0034] Furthermore, step P22 in the embodiments of this application also includes:

[0035] P22-1: Establish a multi-objective optimization function that includes aerodynamic efficiency and mechanical load weights; P22-2: Solve the pitch angle change trajectory that makes the multi-objective optimization function converge through an iterative algorithm; P22-3: Discretize the pitch angle change trajectory into a feedforward pitch angle sequence.

[0036] Specifically, the timing prediction process of the feedforward channel can be further refined by establishing a multi-objective optimization function and solving its convergent pitch angle change trajectory, ultimately generating a feedforward pitch angle sequence.

[0037] First, a multi-objective optimization function is established, incorporating weights for aerodynamic efficiency and mechanical load. Aerodynamic efficiency refers to the efficiency with which a wind turbine converts wind energy into mechanical energy, a crucial indicator of turbine performance. Mechanical load refers to the load borne by the blades and turbine structure, directly impacting the turbine's lifespan and safety. By assigning different weights to aerodynamic efficiency and mechanical load—for example, by introducing weighting factors—the two are unified under a weighted optimization framework, generating the multi-objective optimization function. This weight allocation can be adjusted according to actual operational needs; for instance, at high wind speeds, greater emphasis may be placed on controlling mechanical load, while at low wind speeds, greater emphasis may be placed on improving aerodynamic efficiency.

[0038] Next, an iterative algorithm is used to find the pitch angle variation trajectory that converges the multi-objective optimization function. Iterative algorithms are a commonly used optimization method, gradually approximating the optimal solution through continuous iteration. In this application, particle swarm optimization or deep reinforcement learning methods are preferred to address the high nonlinearity and uncertainty of wind field disturbances. During the iteration process, the pitch angle variation path is continuously adjusted, the optimization function value is calculated, and the convergence condition is determined. When the function converges, a pitch angle variation trajectory balancing power generation efficiency and structural safety is obtained. This trajectory reflects the optimal adjustment method of the pitch angle at different time points to achieve a balance between aerodynamic efficiency and mechanical load.

[0039] Finally, the pitch angle variation trajectory is discretized to generate a feedforward pitch angle sequence. Specifically, based on the sampling period of the control system, the continuous trajectory curve is sampled at equal time intervals to obtain a series of discrete pitch angle command points, thus generating the feedforward pitch angle sequence. Discretization ensures the executability of the trajectory within the digital control system and maintains consistency with the response frequency of the pitch actuator, avoiding delays or oscillations caused by sampling mismatch. The resulting feedforward pitch angle sequence will serve as the basis signal input to the coupled control unit for predictive control, fusing with subsequent feedback compensation signals to provide a basis for generating the optimal pitch angle sequence.

[0040] Furthermore, step P23 in this embodiment of the application also includes:

[0041] P23-1: Extract the power fluctuation feature quantity from the power-temperature correlation matrix; P23-2: Calculate the coupling influence coefficient of structural vibration on power based on the strain space distribution vector; P23-3: Perform weighted correction on the power fluctuation feature quantity based on the coupling influence coefficient to generate dynamic compensation quantity.

[0042] In one possible embodiment of this application, the power deviation compensation process of the feedback channel can be further refined by extracting power fluctuation characteristics, calculating the coupling influence coefficient of structural vibration on power, and generating dynamic compensation based on this information, thereby achieving more precise pitch angle adjustment.

[0043] First, power fluctuation characteristics are extracted from the power-temperature correlation matrix. Power fluctuation characteristics refer to the variations in generator power over different time scales; these characteristics reflect the stability and fluctuation of power output. By analyzing the time series data of power data in the power-temperature correlation matrix, key indicators such as the amplitude, frequency, and energy distribution of power fluctuations over time are identified, allowing the extraction of characteristics related to power fluctuations, i.e., power fluctuation characteristics. These characteristics provide the foundational data for subsequent compensation calculations.

[0044] Next, the coupling influence coefficient of structural vibration on power is calculated based on the strain spatial distribution vector. Specifically, the strain energy density distribution at different azimuth angles of the blade root, characterized by the strain spatial distribution vector, can reveal the vibration characteristics of the blade under non-uniform wind fields. By establishing a structural dynamics model or using modal analysis, the degree of influence of blade vibration on generator power output can be calculated and quantified in the form of a coupling influence coefficient. This coupling influence coefficient reflects the coupling relationship between structural response and power fluctuations; for example, the stronger the blade vibration, the more significant the superimposed fluctuation component in the power output may be.

[0045] Finally, the power fluctuation characteristic quantity is weighted and corrected based on the coupling influence coefficient to generate a dynamic compensation quantity. By combining the coupling influence coefficient with the power fluctuation characteristic quantity, the actual impact of power deviation can be assessed more accurately. For example, the power fluctuation characteristic quantity and the coupling influence coefficient can be weighted to ensure that the compensation quantity reflects the combined effects of aerodynamic disturbances and structural vibrations. Through this weighted correction, the compensation signal can better match the actual operating state of the unit, avoiding control lag or overcompensation caused by relying solely on power deviation. The final generated dynamic compensation quantity will be input to the feedback channel and fused with the feedforward pitch angle sequence to correct control commands in real time, thereby improving the steady-state accuracy and dynamic response capability of the system.

[0046] P30: Drive the pitch actuator according to the optimal pitch angle sequence, and at the same time monitor the blade stress change rate in real time to calculate the aerodynamic load-structure response coupling factor.

[0047] Furthermore, step P30 in this embodiment of the application also includes:

[0048] P31: Obtain the time-domain derivative of the strain spatial distribution vector from the dynamic load spectrum matrix; P32: Perform nonlinear coupling calculation between the time-domain derivative and the turbulent wind speed characteristic component to output the aerodynamic load-structure response coupling factor characterizing the dynamic relationship between aerodynamic load and structural response.

[0049] It should be understood that the pitch control mechanism is driven according to the optimal pitch angle sequence, while the stress change rate of the blades is monitored in real time, and the aerodynamic load-structural response coupling factor is calculated based on this. Through this process, the impact of aerodynamic disturbances on the structure can be dynamically monitored while the pitch angle is adjusted, thus providing a basis for subsequent safety control and fault-tolerant handling.

[0050] Specifically, firstly, the time-domain derivative of the strain spatial distribution vector is obtained from the dynamic load spectrum matrix. The strain spatial distribution vector reflects the spatial distribution of strain energy density at the blade root, while its time-domain derivative further describes the rate of change of strain energy density over time. By numerically differentiating the strain spatial distribution vector over time, the rate of change of strain energy density at the blade root along the time axis can be obtained. This time-domain derivative directly characterizes the dynamic features of the structural response, i.e., the rate of stress change of the blade under wind disturbance. Compared with the static strain distribution, the time-domain derivative better reflects the sensitivity of the structure to transient wind speed impacts or turbulent abrupt changes, and is an important parameter for evaluating structural fatigue loads.

[0051] Next, the time-domain derivative of the strain spatial distribution vector is nonlinearly coupled with the turbulent wind speed characteristic components. These turbulent wind speed characteristic components, obtained through frequency domain eigenvalue decomposition, are key parameters characterizing turbulence intensity. The nonlinear coupling operation is a mathematical process that comprehensively considers the dynamic relationship between aerodynamic loads (caused by turbulent wind speed) and structural responses (reflected by the blade stress change rate). For example, methods such as polynomial fitting, modal superposition, neural network regression, or support vector machines can be used to establish a dynamic mapping relationship between wind speed disturbance characteristics and structural stress change rate, generating an aerodynamic load-structural response coupling factor. This coupling factor, as a comprehensive index, reflects the correlation strength and nonlinearity between aerodynamic disturbance input and structural response output. For instance, when the coupling factor exceeds a preset threshold, it indicates that the dynamic load caused by turbulent impact on the blade structure is approaching a dangerous level, requiring intervention through the control system. By calculating this coupling factor, the structural safety status of the blade can be monitored in real time, providing a scientific basis for subsequent safety threshold judgments and the triggering of backup control paths.

[0052] P40: When the aerodynamic load-structure response coupling factor exceeds the safety threshold, the backup control path is triggered to reconstruct the pitch angle command, and the pitch angle is forcibly adjusted to the preset safety angle according to the reconstructed pitch angle sequence.

[0053] Specifically, when the aerodynamic load-structural response coupling factor exceeds a preset safety threshold, the system will quickly trigger a backup control path to reconstruct the pitch angle command and forcibly adjust the pitch angle to a preset safe angle. The aerodynamic load-structural response coupling factor, as a real-time operational status monitoring indicator, is used to quantify the dynamic coupling relationship between turbulent wind speed disturbances and blade structural stress changes. When this factor exceeds the preset safety threshold, it indicates that the coupling effect between the aerodynamic load and structural response has reached a dangerous level that may cause fatigue damage or transient failure. At this point, the control system should immediately enter the backup control mode.

[0054] In the backup control path, the existing optimal pitch angle sequence is first reconstructed to generate new pitch angle commands. For example, a fast correction algorithm or a command reconstruction method based on constraint optimization can be used to shift the control objective from efficiency priority to safety priority, focusing on constraining the stress change rate at the blade root and the thermal load level of the pitch bearing, thereby forming a reconstructed pitch angle sequence with protective functions. Subsequently, based on the reconstructed pitch angle sequence, the blade angle is forcibly adjusted. The forced adjustment process needs to bypass some conventional optimization strategies to ensure a rapid reduction in blade stress and structural vibration levels within a short period of time, thereby preventing the unit from entering a dangerous state. Finally, the pitch angle is adjusted to a preset safe angle range to ensure that the blades are under a tolerable stress state under extreme wind conditions, while maintaining the basic operational stability of the unit.

[0055] In practical applications, the activation of backup control paths and the forced adjustment of pitch angle require a highly reliable and fast-response control system, including high-performance sensors, fast-processing controllers, and precise actuators. Furthermore, the setting of preset safety angles and safety thresholds requires precise calculation and verification based on detailed unit design parameters and safety standards to ensure effective protection of the unit under various operating conditions.

[0056] Furthermore, in this embodiment of the application, step P40, which triggers the backup control path to reconstruct the pitch angle command, also includes:

[0057] P41: Calculate the critical pitch angle range under the current wind conditions based on the turbulent wind speed characteristic components in the dynamic load spectrum matrix; P42: Determine the structural safety constraint boundary according to the extreme value distribution of the strain space distribution vector; P43: Generate a reconstructed pitch angle sequence that satisfies multi-objective optimization within the intersection of the critical pitch angle range and the structural safety constraint boundary.

[0058] Optionally, the process of triggering the backup control path to reconfigure the pitch angle command can be further refined when the aerodynamic load-structure response coupling factor exceeds the safety threshold.

[0059] First, based on the turbulent wind speed characteristic components in the dynamic load spectrum matrix, the critical pitch angle range under the current wind conditions is calculated. The turbulent wind speed characteristic components are obtained through frequency domain eigenvalue decomposition and are key parameters characterizing turbulence intensity. The critical pitch angle range refers to the pitch angle interval within which the blades can operate safely under the current wind conditions. Therefore, by analyzing the statistical values, spectral energy distribution, and intensity components of the turbulent wind speed characteristics, and using the unit's aerodynamic performance model, the upper and lower limits of the pitch angle that can maintain stable operation of the unit under these wind conditions can be solved, thus obtaining the critical pitch angle range. By calculating this range, the system can determine the maximum and minimum pitch angles that the blades can withstand under extreme wind conditions, thereby providing a safety boundary for subsequent pitch angle adjustments.

[0060] Next, the structural safety constraint boundary is determined based on the extreme value distribution of the strain spatial distribution vector. The strain spatial distribution vector reflects the spatial distribution of strain energy density at the blade root, while its extreme value distribution further describes the maximum stress at different locations on the blade. Therefore, by extracting the extreme values ​​of each partition of the strain spatial distribution vector at the blade root, the location and amplitude of the maximum stress response of the blade structure under the current operating conditions can be obtained. Combined with the material strength limit and fatigue damage accumulation criterion, structural safety boundary conditions are established, which represent the maximum range within which the pitch angle can be adjusted under the premise of blade structural safety. By determining this boundary, the system can ensure that adjusting the pitch angle does not lead to excessive stress on the blade structure, thereby avoiding potential structural damage.

[0061] Finally, within the intersection of the critical pitch angle range and the structural safety constraint boundary, a reconfigurable pitch angle sequence satisfying multi-objective optimization is generated. For example, power generation stability, mechanical load suppression, and actuator response characteristics can be comprehensively optimized as multi-objective functions. Under the premise of ensuring the pitch angle remains within the safe intersection range, a reconfigurable pitch angle sequence is generated through iterative solutions or constraint optimization algorithms. This ensures that, under safe conditions, pitch angle adjustments can maximize power generation efficiency and reduce load. This reconfigurable pitch angle sequence can provide the control system with a control path that balances aerodynamic performance and structural safety in emergency situations, thereby ensuring that the wind turbine generator maintains a safe and stable operating state under extreme turbulence or sudden load changes.

[0062] Furthermore, according to the reconstructed pitch angle sequence, the pitch angle is forcibly adjusted to a preset safe angle. In this embodiment, step P40 further includes:

[0063] P44: When the coupling factor is in the first-level over-limit range, the pitch rate is limited according to the gradient ratio of the reconstructed pitch angle sequence; P45: When the coupling factor is in the second-level over-limit range, the pitch rate is directly switched to the limit safety angle in the reconstructed pitch angle sequence; P46: The adjusted strain energy density change rate is fed back to the dynamic load spectrum matrix in real time.

[0064] Specifically, a graded control strategy is adopted based on the level of exceedance of the aerodynamic load-structural response coupling factor to achieve differentiated responses to different levels of hazardous conditions, thereby improving the flexibility and effectiveness of backup control paths.

[0065] Specifically, when the coupling factor is within the first-level out-of-limit range, the system will limit the pitch rate according to the gradient ratio of the reconstructed pitch angle sequence. The first-level out-of-limit range refers to the situation where the coupling factor exceeds the safety threshold but has not yet reached an extremely dangerous state. In this case, the system calculates the gradient of the reconstructed pitch angle sequence and sets a rate upper limit proportional to the gradient value to control the operating speed of the pitch actuator, avoiding structural shocks or other potential risks caused by excessively rapid adjustments. This method can ensure that the pitch angle gradually transitions to the safe range while avoiding mechanical shocks or actuator fatigue caused by excessively rapid adjustments, thus achieving a balance between safety and smooth operation.

[0066] When the coupling factor is in the second-order over-limit range, the system will directly switch to the ultimate safe angle in the reconstructed pitch angle sequence. The second-order over-limit range refers to the coupling factor exceeding the first-order over-limit range, placing the unit in a highly dangerous operating condition. Delayed adjustment could lead to structural failure or stall risk. Therefore, a direct switch is used to quickly place the blade angle at the pre-set ultimate safe position to effectively reduce aerodynamic loads and structural response in the shortest possible time, thereby preventing fault propagation. The ultimate safe angle is pre-set and can minimize loads under extreme conditions, protecting the unit from damage.

[0067] Finally, the strain energy density at the blade root is monitored in real time and its rate of change is calculated. The rate of change of strain energy density is fed back to the dynamic load spectrum matrix for updating, so as to ensure that the dynamic load spectrum matrix continuously reflects the latest structural response state and provides accurate input data for subsequent control links, thereby realizing closed-loop optimization and adaptive control, ensuring effective protection of the unit under various operating conditions, and thus maintaining the long-term stability and adaptive capability of the system.

[0068] In summary, the embodiments of this application have at least the following technical effects:

[0069] This application utilizes a multi-source heterogeneous sensor array to collect turbulent wind speed, strain energy density, power, and temperature data in real time and constructs a dynamic load spectrum matrix to achieve a comprehensive characterization of aerodynamic disturbances and structural states. Nonlinear optimization is performed through a coupled control unit, combined with feedforward prediction and feedback compensation mechanisms to generate the optimal pitch angle sequence, thereby improving the real-time performance and accuracy of pitch angle adjustment. By introducing an aerodynamic load-structural response coupling factor as a safety criterion, the nonlinear relationship between aerodynamic disturbances and structural response can be dynamically monitored and risks can be identified in a timely manner. When the coupling factor exceeds the limit, a backup control path is triggered to reconstruct the pitch angle command and forcibly adjust it to a preset safe angle, achieving active protection and safe operation under extreme wind conditions. This results in a dynamic balance between power generation efficiency and structural safety in complex turbulent environments, improving the operational reliability and service life of wind turbine generators.

[0070] The technology achieves a dynamic balance between power generation efficiency and structural safety in complex turbulent environments by integrating multi-source data and nonlinear optimization control, and introducing a backup control path for over-limit triggering.

[0071] Example 2, based on the same inventive concept as the intelligent pitch control method for wind turbine generators in the foregoing examples, such as... Figure 2 As shown, this application provides an intelligent pitch control system for wind turbine generator sets. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0072] The multi-source data acquisition module 11 is used to collect turbulent wind speed, blade root strain energy density, generator instantaneous power and pitch bearing temperature in real time through a multi-source heterogeneous sensor array, and construct a dynamic load spectrum matrix.

[0073] The nonlinear optimization module 12 is used to input the dynamic load spectrum matrix into the coupling control unit for nonlinear optimization to generate the optimal pitch angle sequence.

[0074] The dynamic monitoring factor calculation module 13 is used to drive the pitch actuator according to the optimal pitch angle sequence, while monitoring the blade stress change rate in real time and calculating the aerodynamic load-structure response coupling factor.

[0075] The pitch angle reconstruction module 14 is used to trigger the backup control path to reconstruct the pitch angle command when the aerodynamic load-structure response coupling factor exceeds the safety threshold, and to forcibly adjust the pitch angle to a preset safety angle according to the reconstructed pitch angle sequence.

[0076] Furthermore, the multi-source data acquisition module 11 is also used to perform the following steps:

[0077] The collected turbulent wind speed is decomposed in the frequency domain to obtain the characteristic components characterizing the turbulence intensity; the strain energy density at the blade root is sampled in circumferential angular partitions to generate a strain spatial distribution vector; the instantaneous power of the generator and the temperature of the pitch bearing are correlated in a time series to form a power-temperature correlation matrix; the characteristic components, the strain spatial distribution vector and the power-temperature correlation matrix are fused to generate a dynamic load spectrum matrix.

[0078] Furthermore, the nonlinear optimization module 12 is also used to perform the following steps:

[0079] The coupling control unit includes a feedforward channel and a feedback channel; the feedforward channel performs time-series prediction of the turbulent wind speed characteristics in the dynamic load spectrum matrix to generate a feedforward pitch angle sequence; the feedback channel dynamically generates a compensation amount based on the generator power deviation in the dynamic load spectrum matrix; the feedforward pitch angle sequence and the compensation amount are fused to output the optimal pitch angle sequence.

[0080] Furthermore, the nonlinear optimization module 12 is also used to perform the following steps:

[0081] A multi-objective optimization function with aerodynamic efficiency and mechanical load weights is established; the pitch angle change trajectory that makes the multi-objective optimization function converge is solved by an iterative algorithm; and the pitch angle change trajectory is discretized into a feedforward pitch angle sequence.

[0082] Furthermore, the nonlinear optimization module 12 is also used to perform the following steps:

[0083] Extract the power fluctuation feature from the power-temperature correlation matrix; calculate the coupling influence coefficient of structural vibration on power based on the strain space distribution vector; and perform weighted correction on the power fluctuation feature based on the coupling influence coefficient to generate dynamic compensation.

[0084] Furthermore, the dynamic monitoring factor calculation module 13 is also used to perform the following steps:

[0085] The time-domain derivative of the strain spatial distribution vector is obtained from the dynamic load spectrum matrix; the time-domain derivative is nonlinearly coupled with the turbulent wind speed characteristic component to output the aerodynamic load-structure response coupling factor, which characterizes the dynamic relationship between aerodynamic load and structural response.

[0086] Furthermore, the pitch angle reconstruction module 14 is also used to perform the following steps:

[0087] Based on the turbulent wind speed characteristic components in the dynamic load spectrum matrix, the critical pitch angle range under the current wind conditions is calculated; according to the extreme value distribution of the strain space distribution vector, the structural safety constraint boundary is determined; within the intersection of the critical pitch angle range and the structural safety constraint boundary, a reconstructed pitch angle sequence that satisfies multi-objective optimization is generated.

[0088] Furthermore, the pitch angle reconstruction module 14 is also used to perform the following steps:

[0089] When the coupling factor is in the first-level over-limit range, the pitch rate is limited according to the gradient ratio of the reconstructed pitch angle sequence; when the coupling factor is in the second-level over-limit range, the pitch rate is directly switched to the limit safety angle in the reconstructed pitch angle sequence; the adjusted strain energy density change rate is fed back to the dynamic load spectrum matrix in real time.

[0090] Example 3: Exemplary electronic device.

[0091] The following is for reference. Figure 3 The present application describes the electronic device according to its embodiments.

[0092] Based on the same inventive concept as the intelligent pitch control method for wind turbine generator sets in the foregoing embodiments, this application also provides an intelligent pitch control system for wind turbine generator sets, including: a processor coupled to a memory for storing a program, which, when executed by the processor, causes the system to perform the steps of the method described in Embodiment 1.

[0093] The electronic device 300 includes a processor 302, a communication interface 303, and a memory 301. Optionally, the electronic device 300 may also include a bus architecture 304. The communication interface 303, processor 302, and memory 301 can be interconnected via the bus architecture 304; the bus architecture 304 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus architecture 304 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0094] Processor 302 may be a CPU, microprocessor, ASIC, or one or more integrated circuits used to control the execution of programs according to the present application.

[0095] Communication interface 303 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), wired access network, etc.

[0096] Memory 301 can be ROM or other types of static storage devices capable of storing static information and instructions, RAM or other types of dynamic storage devices capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory can exist independently and be connected to the processor via bus architecture 304. Memory can also be integrated with the processor.

[0097] The memory 301 stores computer execution instructions for implementing the scheme of this application, and the processor 302 controls the execution. The processor 302 executes the computer execution instructions stored in the memory 301, thereby implementing the intelligent pitch control method for wind turbine generators provided in the above embodiments of this application.

[0098] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0099] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0100] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A smart pitch control method for wind turbine generator sets, characterized in that, The method includes: A dynamic load spectrum matrix is ​​constructed by synchronously collecting turbulent wind speed, blade root strain energy density, generator instantaneous power, and pitch bearing temperature in real time using a multi-source heterogeneous sensor array. The dynamic load spectrum matrix is ​​input into the coupling control unit for nonlinear optimization to generate the optimal pitch angle sequence. Drive the pitch actuator according to the optimal pitch angle sequence, and simultaneously monitor the blade stress change rate in real time to calculate the aerodynamic load-structural response coupling factor. When the aerodynamic load-structure response coupling factor exceeds the safety threshold, the backup control path is triggered to reconstruct the pitch angle command, and the pitch angle is forcibly adjusted to the preset safety angle according to the reconstructed pitch angle sequence.

2. The intelligent pitch control method for wind turbine generator sets as described in claim 1, characterized in that, The construction of the dynamic load spectrum matrix includes: The collected turbulent wind speeds are decomposed in the frequency domain to obtain characteristic components representing the intensity of turbulence. The strain energy density at the blade root is sampled in circumferential angular partitions to generate a strain spatial distribution vector. The instantaneous power of the generator and the temperature of the pitch bearing are correlated in a time series to form a power-temperature correlation matrix; By fusing the aforementioned characteristic components, strain spatial distribution vector, and power-temperature correlation matrix, a dynamic load spectrum matrix is ​​generated.

3. The intelligent pitch control method for wind turbine generator sets as described in claim 2, wherein the dynamic load spectrum matrix is ​​input to the coupling control unit for nonlinear optimization to generate the optimal pitch angle sequence, comprising: The coupling control unit includes a feedforward channel and a feedback channel; The feedforward channel is used to perform time-series prediction of the turbulent wind speed characteristics in the dynamic load spectrum matrix to generate a feedforward blade pitch angle sequence. The compensation amount is dynamically generated through the feedback channel based on the generator power deviation in the dynamic load spectrum matrix. By combining the feedforward pitch angle sequence with the compensation amount, the optimal pitch angle sequence is output.

4. The intelligent pitch control method for wind turbine generator sets as described in claim 3, characterized in that, The feedforward channel is used to perform time-series prediction of the turbulent wind speed characteristics in the dynamic load spectrum matrix, generating a feedforward blade pitch angle sequence, including: Establish a multi-objective optimization function that includes weights for aerodynamic efficiency and mechanical load; The pitch angle variation trajectory that makes the multi-objective optimization function converge is obtained by iterative algorithm; The pitch angle variation trajectory is discretized into a feedforward pitch angle sequence.

5. The intelligent pitch control method for wind turbine generator sets as described in claim 4, characterized in that, The compensation amount is dynamically generated through the feedback channel based on the generator power deviation in the dynamic load spectrum matrix, including: Extract the power fluctuation feature from the power-temperature correlation matrix; The coupling influence coefficient of structural vibration on power is calculated based on the strain space distribution vector. The power fluctuation characteristic quantity is weighted and corrected based on the coupling influence coefficient to generate a dynamic compensation quantity.

6. The intelligent pitch control method for wind turbine generator sets as described in claim 2, characterized in that, Real-time monitoring of blade stress change rate and calculation of aerodynamic load-structural response coupling factor, including: Obtain the time-domain derivative of the strain space distribution vector from the dynamic load spectrum matrix; The time-domain derivative is nonlinearly coupled with the characteristic components of turbulent wind speed to output the aerodynamic load-structure response coupling factor, which characterizes the dynamic relationship between aerodynamic load and structural response.

7. The intelligent pitch control method for wind turbine generator sets as described in claim 2, characterized in that, When the aerodynamic load-structure response coupling factor exceeds a safety threshold, a backup control path reconfiguration pitch angle command is triggered, which also includes: Based on the turbulent wind speed characteristic components in the dynamic load spectrum matrix, the critical pitch angle range under the current wind conditions is calculated. The structural safety constraint boundary is determined based on the extreme value distribution of the strain space distribution vector. Within the intersection of the critical pitch angle range and the structural safety constraint boundary, a reconfigured pitch angle sequence that satisfies multi-objective optimization is generated.

8. The intelligent pitch control method for wind turbine generator sets as described in claim 7, characterized in that, Forcefully adjust the pitch angle to a preset safe angle based on the reconstructed pitch angle sequence, including: When the coupling factor is in the first-level out-of-limit range, the pitch rate is limited according to the gradient ratio of the reconstructed pitch angle sequence; When the coupling factor is in the second-order over-limit range, directly switch to the limit safety angle in the reconstructed pitch angle sequence; The adjusted strain energy density change rate is fed back to the dynamic load spectrum matrix in real time.

9. An intelligent pitch control system for wind turbine generator sets, characterized in that, The system includes: The multi-source data acquisition module is used to collect turbulent wind speed, blade root strain energy density, generator instantaneous power and pitch bearing temperature in real time through a multi-source heterogeneous sensor array, and construct a dynamic load spectrum matrix. The nonlinear optimization module is used to input the dynamic load spectrum matrix into the coupling control unit for nonlinear optimization to generate the optimal pitch angle sequence. The dynamic monitoring factor calculation module is used to drive the pitch actuator according to the optimal pitch angle sequence, while monitoring the blade stress change rate in real time and calculating the aerodynamic load-structure response coupling factor. The pitch angle reconstruction module is used to trigger the backup control path to reconstruct the pitch angle command when the aerodynamic load-structure response coupling factor exceeds the safety threshold, and to forcibly adjust the pitch angle to a preset safety angle according to the reconstructed pitch angle sequence.

10. An electronic device, characterized in that, include: A processor coupled to a memory for storing a program that, when executed by the processor, causes the system to perform the steps of the method as claimed in any one of claims 1 to 8.

Images