Yaw error cooperative compensation method and system for a variable pitch system

Abstract

This disclosure provides a method and system for collaborative compensation of yaw error in a pitch control system. It acquires the blade's bending moment matrix, azimuth angle, and macroscopic yaw error, and uses asymmetric load decoupling and quantization techniques to convert the blade load into fixed d-axis tilting moment and q-axis yaw moment components. Subsequently, these moment components are subjected to dual-channel PI control to generate d-axis and q-axis pitch compensation commands, and the yaw feedforward signal is cleverly extracted from the q-axis integral term. Finally, based on the azimuth angle, the d-axis and q-axis pitch compensation commands are fused with the collective pitch command to obtain independent pitch commands for each blade. These commands are then combined with the yaw feedforward signal and macroscopic yaw error for collaborative control and execution, thereby achieving deep collaboration between the pitch control and yaw systems. This approach more accurately reflects the rotor's stress state and enables refined compensation, effectively reducing fatigue loads on the blades and tower base.

Description

Technical Field

[0001] This invention relates to the field of intelligent compensation, and in particular to a method and system for collaborative compensation of yaw error in a pitch system. Background Technology

[0002] During operation, wind turbines often exhibit yaw errors due to environmental factors such as wind speed and direction. This yaw error prevents the rotor from being perfectly aligned with the oncoming wind, leading to a series of negative consequences. First, yaw error subjects the rotor blades to asymmetric aerodynamic loads, significantly increasing cyclic loads such as blade root bending moment and tower base load. Over time, this accelerates structural fatigue damage, shortens the lifespan of critical components, and increases maintenance costs. Second, asymmetric loads can induce vibrations, affecting the stability of the turbine's operation. More importantly, yaw error directly impacts wind energy capture efficiency, reducing power generation and thus weakening the economic benefits of wind farms. Therefore, to improve the reliability, power generation efficiency, and extend the lifespan of wind turbines, it is crucial to develop a coordinated control scheme that effectively compensates for yaw errors and the resulting asymmetric loads.

[0003] Existing pitch control systems mostly rely on macroscopic yaw error compensation schemes measured by anemometers and wind vanes on top of the nacelle. These traditional methods typically adjust the yaw angle of the entire nacelle to align the rotor with the average wind direction. However, this compensation strategy based on macroscopic yaw error has significant technical limitations. On the one hand, relying solely on a scalarized macroscopic yaw error cannot accurately capture the true wind field distribution on the rotor plane and the resulting asymmetric load state, nor can it provide fine-grained compensation for these asymmetric loads. On the other hand, the sensors on top of the nacelle themselves have inertia, and their measurement data often lags behind changes in the actual wind field. Under conditions of rapid wind direction changes, compensation commands calculated based on lag information may not match actual load changes, and may even produce a "reverse compensation" effect in certain phases, exacerbating blade loads. Furthermore, the traditional control objective is often to eliminate yaw error, but zero yaw error does not necessarily mean minimal asymmetric loads. Sometimes, maintaining zero yaw error may actually increase asymmetric loads on the blades, creating a conflict between the control objective and the actual physical problem. Traditional solutions face challenges in accurately quantifying and effectively compensating for asymmetric loads, and cannot fully utilize the sensor information of the blades themselves to directly reflect and compensate for asymmetric loads caused by yaw errors.

[0004] Therefore, an optimized method for collaborative compensation of yaw error in pitch control systems is needed. Summary of the Invention

[0005] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide a method and system for collaborative compensation of yaw error in a pitch control system. It acquires the blade's bending moment matrix, azimuth angle, and macroscopic yaw error, and uses asymmetric load decoupling and quantization technology to convert the blade load into fixed d-axis tilting moment and q-axis yaw moment components. Subsequently, these moment components are subjected to dual-channel PI control to generate d-axis and q-axis pitch compensation commands, and the yaw feedforward signal is cleverly extracted from the q-axis integral term. Finally, based on the azimuth angle, the d-axis and q-axis pitch compensation commands are fused with the collective pitch command to obtain independent pitch commands for each blade. These commands are then combined with the yaw feedforward signal and macroscopic yaw error for collaborative control and execution, thereby achieving deep collaboration between the pitch control and yaw systems. This approach more accurately reflects the rotor's stress state and enables refined compensation, effectively reducing fatigue loads on the blades and tower base.

[0006] According to one aspect of this application, a method for coordinated compensation of yaw error in a pitch control system is provided, comprising: Obtain the bending moment matrix, azimuth angle, and macroscopic yaw error of the first to third blades; Asymmetric load decoupling and quantization are performed on the bending moment matrix and azimuth angle of the first to third blades to obtain the fixed d-axis tilting moment component and the fixed q-axis yaw moment component. Dual-channel PI control is performed on the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to obtain the d-axis pitch compensation command, the q-axis pitch compensation command and the yaw feedforward signal; Based on the azimuth angle, the d-axis pitch compensation command, the q-axis pitch compensation command, and the collective pitch command are fused to obtain the independent pitch commands for the first to third blades. Coordinated control and execution are achieved based on the independent pitch control commands of the first to third blades, the yaw feedforward signal, and the macroscopic yaw error.

[0007] According to another aspect of this application, a pitch system yaw error collaborative compensation system is provided, comprising: The data acquisition module is used to acquire the bending moment matrix, azimuth angle, and macroscopic yaw error of the first to third blades; The asymmetric load decoupling and quantization module is used to perform asymmetric load decoupling and quantization on the bending moment matrix and azimuth angle of the first to third blades to obtain the fixed d-axis tilting moment component and the fixed q-axis yaw moment component. The dual-channel PI control module is used to perform dual-channel PI control on the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to obtain the d-axis pitch compensation command, the q-axis pitch compensation command and the yaw feedforward signal. The command fusion module is used to fuse the d-axis pitch compensation command, q-axis pitch compensation command and collective pitch command based on the azimuth angle to obtain the independent pitch commands for the first to third blades. The collaborative control and execution module is used for collaborative control and execution based on the independent pitch control commands of the first to third blades, the yaw feedforward signal, and the macroscopic yaw error.

[0008] Compared with existing technologies, this application provides a method and system for collaborative compensation of yaw error in a pitch control system. It acquires the blade's bending moment matrix, azimuth angle, and macroscopic yaw error, and uses asymmetric load decoupling and quantization techniques to convert the blade load into fixed d-axis tilting moment and q-axis yaw moment components. Subsequently, these moment components are subjected to dual-channel PI control to generate d-axis and q-axis pitch compensation commands, and the yaw feedforward signal is cleverly extracted from the q-axis integral term. Finally, based on the azimuth angle, the d-axis and q-axis pitch compensation commands are fused with the collective pitch command to obtain independent pitch commands for each blade. These commands are then combined with the yaw feedforward signal and macroscopic yaw error for collaborative control and execution, thereby achieving deep collaboration between the pitch control and yaw systems. This approach more accurately reflects the rotor's stress state and allows for refined compensation, effectively reducing fatigue loads on the blades and tower base. Attached Figure Description

[0009] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 This is a flowchart of the pitch system yaw error collaborative compensation method according to an embodiment of this application; Figure 2 This is a schematic diagram of the data flow of the pitch system yaw error collaborative compensation method according to an embodiment of this application; Figure 3 This is a block diagram of a pitch system yaw error collaborative compensation system according to an embodiment of this application. Detailed Implementation

[0011] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0012] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not specifically singular and may include plural forms. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0013] While this application makes various references to certain modules of the systems according to embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The modules described are merely illustrative, and different aspects of the systems and methods may use different modules.

[0014] Flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0015] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0016] In the technical solution of this application, a collaborative compensation method for yaw error of a pitch system is proposed. Figure 1 This is a flowchart of a pitch system yaw error collaborative compensation method according to an embodiment of this application. Figure 2 This is a system architecture diagram of the pitch system yaw error collaborative compensation method according to an embodiment of this application. Figure 1 and Figure 2 As shown, the pitch system yaw error collaborative compensation method according to an embodiment of this application includes the following steps: S1, obtaining the bending moment matrix, azimuth angle, and macroscopic yaw error of the first to third blades; S2, performing asymmetric load decoupling and quantization on the bending moment matrix and azimuth angle of the first to third blades to obtain a fixed d-axis tilt moment component and a fixed q-axis yaw moment component; S3, performing dual-channel PI control on the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to obtain d-axis pitch compensation command, q-axis pitch compensation command, and yaw feedforward signal; S4, based on the azimuth angle, performing command fusion on the d-axis pitch compensation command, q-axis pitch compensation command, and collective pitch command to obtain independent pitch commands for the first to third blades; S5, performing collaborative control and execution based on the independent pitch commands of the first to third blades, the yaw feedforward signal, and the macroscopic yaw error.

[0017] Specifically, S1 involves acquiring the bending moment matrix, azimuth angle, and macroscopic yaw error of the first to third blades. It should be understood that asymmetric loads are the direct cause of increased fatigue damage and reduced power generation efficiency in the unit, and yaw error is a significant contributing factor. Therefore, only by accurately acquiring bending moment data that directly reflects the load state, azimuth angle information necessary for load decoupling, and macroscopic yaw error characterizing the overall wind conditions of the unit, can subsequent load decoupling, controller design, and the generation of coordinated compensation commands have practical significance and effectiveness, thereby ensuring that the entire system can respond quickly and accurately to real operating conditions.

[0018] Among them, the bending moment matrix is ​​a vector or matrix whose elements are the bending moment values ​​measured by the sensors at the roots of the first, second, and third blades at the same moment. It is the core basis for evaluating the asymmetric load of the wind turbine. The azimuth angle refers to the instantaneous angular position of the wind turbine during rotation. The 0-degree position is usually defined as the blades pointing vertically upward. It determines the relative position of each blade in a fixed coordinate system. The macroscopic yaw error refers to the angular deviation between the nacelle direction of the wind turbine and the direction of the incoming wind. It macroscopically reflects the wind state of the unit and is one of the fundamental causes of asymmetric load.

[0019] In practice, the acquisition of the bending moment matrix and azimuth angle of the first to third blades typically relies on a hardware sensor system installed on the wind turbine. Specifically, strain sensors (such as fiber optic gratings or strain gauges) can be installed at the root of each blade. These sensors can monitor the deformation of the blades caused by wind in real time and convert it into electrical signals. After calculation by the data acquisition and processing system, the real-time bending moment value of each blade in the flapping or oscillating direction can be obtained. These values ​​together constitute the bending moment matrix. At the same time, a rotary encoder or similar position sensor installed on the wind turbine main shaft or generator rotor will continuously output a high-precision angle signal, which is the real-time azimuth angle of the wind turbine. This provides a basis for determining the instantaneous spatial position of each blade during the rotation cycle. Secondly, obtaining the macroscopic yaw error is not a simple direct measurement, but a physical calculation process. Specifically, firstly, the difference between the wind direction angle and the nacelle yaw angle is calculated as the raw error; where the wind direction angle defines the direction of wind flow on a horizontal plane. This is an absolute angular quantity, requiring a fixed reference benchmark. In the wind power industry, this reference benchmark is usually geographic north. Therefore, the wind direction angle can be precisely described as: the angle traversed by rotating clockwise or counterclockwise from geographic north as 0 degrees to the direction of the wind's origin; for example, if the wind is blowing from due east, and the system defines clockwise as positive, then the wind direction angle is 90 degrees; if the wind is blowing from due south, the wind direction angle is 180 degrees. This angle is a key parameter describing the wind field environment, directly determining the specific direction in which the wind turbine can maximize energy capture; the wind direction angle is obtained through a wind vane installed on top of the wind turbine nacelle. The wind vane is usually located at the rear of the nacelle, forming a weather station together with the anemometer. The nacelle yaw angle defines the orientation of the wind turbine nacelle itself on the horizontal plane. Like the wind direction angle, it is an absolute angular quantity and uses the same reference point, geographic north. Therefore, the nacelle yaw angle can be precisely described as the angle traversed by rotating clockwise or counterclockwise from geographic north (0 degrees) to the direction normal to the rotor's plane of rotation. This angle represents the turbine's current attitude and is a variable that the control system needs to actively adjust. The nacelle yaw angle is obtained through a yaw position sensor installed in the turbine's yaw system, which directly measures the angle of rotation of the nacelle relative to the stationary tower. Then, the original error is angle-normalized using the following formula to obtain the macroscopic yaw error: , in, This is the original error. For finding the remainder function.

[0020] Specifically, in step S2, asymmetric load decoupling and quantization are performed on the bending moment matrix and azimuth angle of the first to third blades to obtain a fixed d-axis tilting moment component and a fixed q-axis yaw moment component. It should be understood that the bending moment signals directly measured from the three rotating blades are highly coupled and rapidly time-varying with the azimuth angle, making them difficult to use directly as input for stable control. These measurements in the rotating coordinate system need to be converted into torque components with clear physical meaning, relative stability, and mutual orthogonality in the fixed coordinate system. Therefore, in the technical solution of this application, asymmetric load decoupling and quantization are performed on the bending moment matrix and azimuth angle of the first to third blades to simplify the complex asymmetric load problem into the problem of adjusting two independent DC components (tilting moment and yaw moment). This greatly simplifies the controller design and improves the robustness of the control system.

[0021] In practice, the first step is to determine the Coleman transformation matrix based on the azimuth angle. Since the coordinate system of the wind turbine blades rotates along with the wind turbine, its instantaneous direction relative to the fixed ground coordinate system (usually defined as the dq coordinate system) is constantly changing, and this change is quantitatively described by the azimuth angle. Therefore, in order to accurately project the bending moment on the rotating blades onto the fixed dq axis at any given time, the transformation matrix itself must be a function of the azimuth angle, capable of dynamically adjusting its internal parameters according to the real-time azimuth angle. During this process, an independent module in the control system continuously receives real-time azimuth angle measurements from sensors such as the spindle encoder. For each sampling period, this module immediately performs a matrix construction operation to generate the Coleman transformation matrix corresponding to the current moment. Next, based on the Coleman transformation matrix, the bending moment matrices of the first to third blades are subjected to load projection from a rotating coordinate system to a fixed coordinate system to obtain the unscaled dq-axis moment component vectors. It should be understood that the directly measured bending moment values ​​of the three blades are time-varying AC quantities in their respective rotating coordinate systems. They are coupled to each other, and their magnitudes are not only related to the asymmetric load but also closely related to the instantaneous rotational position (azimuth angle) of the blades. This type of data is not convenient for direct use in designing stable, decoupled controllers. Therefore, in the technical solution of this application, the bending moment matrices of the first to third blades are subjected to load projection from a rotating coordinate system to a fixed coordinate system (i.e., the dq coordinate system) to project these complex time-varying signals into a fixed coordinate system (i.e., the dq coordinate system) that is fixed to the nacelle. Load projection is a process that combines geometric and physical meaning. It projects the bending moment vector, representing the total load state, in three-dimensional space (defined by the three blade axes) onto another orthogonal three-dimensional space (defined by the d, q, and 0 axes) by left-multiplying it by a transformation matrix, obtaining the components on each coordinate axis in that space. In this way, the asymmetric load information contained in the three AC signals can be extracted and transformed into two relatively stable DC components (d-axis and q-axis components), paving the way for subsequent applications of modern control theories such as PI control. Specifically, the bending moment matrix of the first to third blades is rotated from a coordinate system to a fixed coordinate system using the following formula: , in, Here is the Coleman transformation matrix. Here are the bending moment matrices for the first to third blades; Furthermore, the unscaled dq-axis moment component vector is extracted and scaled for power invariance to obtain the fixed d-axis tilting moment component and the fixed q-axis yaw moment component. It should be understood that during wind turbine operation, the bending moment borne by the blades is constantly changing, and its direction changes with the blade rotation. This load information in a rotating coordinate system is inconvenient for direct yaw error compensation control. By introducing the concept of a fixed dq-axis, the complex rotating load can be decoupled into two components with clear physical meaning: the d-axis tilting moment component mainly represents the moment acting on the hub, attempting to make it tilt up and down, mainly caused by wind shear and yaw error; the q-axis yaw moment component represents the moment attempting to make it yaw left and right, mainly caused by yaw error. Transforming these torque components to a fixed coordinate system and performing power invariance scaling not only simplifies the controller design, enabling it to compensate for fixed components, but also ensures that the system power is conserved before and after the coordinate transformation, maintaining physical accuracy. This provides precise and standardized input signals for subsequent dual-channel PI control, thereby achieving effective coordinated compensation for yaw errors.

[0022] Specifically, in S3, dual-channel PI control is applied to the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to obtain the d-axis pitch compensation command, the q-axis pitch compensation command, and the yaw feedforward signal. It should be understood that during wind turbine operation, undesirable tilt and yaw moments are generated due to factors such as wind shear, tower shadow effect, and actual yaw error. If these moment components are not controlled, they will lead to uneven blade load, affecting the stable operation of the generator set and potentially shortening component life. Therefore, in the technical solution of this application, dual-channel PI control is applied to the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to actively and effectively compensate for the fixed d-axis tilt moment component and the fixed q-axis yaw moment component borne by the wind turbine, thereby generating corresponding pitch compensation commands and yaw feedforward signals to reduce system yaw error and improve operational stability. Specifically, by applying dual-channel PI control to these moment components, real-time and precise adjustments can be made to the deviations generated by the d-axis tilt moment and the q-axis yaw moment, respectively. In this process, the PI controller, with its superior ability to eliminate steady-state errors and improve system dynamic response, ensures that the generated control commands can respond quickly to load changes and ultimately bring the system to the desired stable state. Simultaneously, by extracting the yaw feedforward signal from the q-axis integral term through nonlinear mapping, an advance compensation amount based on historical error accumulation can be provided to the yaw system, further improving the response speed and accuracy of yaw control and achieving more efficient collaborative compensation.

[0023] In practice, firstly, the fixed d-axis tilting moment component and the fixed q-axis yaw moment component are input as control errors into the PI controller to obtain the d-axis proportional term, d-axis integral term, q-axis proportional term, and q-axis integral term. For the d-axis PI controller, the input is... The output is the d-axis scaling term. and d-axis integral term d-axis scale term With current error Proportional, d-axis integral term That is the error The accumulation over time; for a q-axis PI controller, its input is... The output is the q-axis scaling term. and q-axis integral term q-axis scaling term With current error Proportional, q-axis integral term That is the error Accumulation over time; Next, the d-axis proportional, d-axis integral, q-axis proportional, and q-axis integral terms are synthesized and output with anti-saturation processing to obtain the d-axis pitch compensation command and the q-axis pitch compensation command. Specifically, the d-axis pitch compensation command is generated by summing the d-axis proportional and d-axis integral terms; this process is expressed by the formula: , in, This is the d-axis pitch compensation command; Similarly, the q-axis pitch compensation command is generated by summing the q-axis proportional term and the q-axis integral term; this process can be expressed by the formula: , in, This is the q-axis pitch compensation command.

[0024] Specifically, an anti-saturation mechanism is introduced during instruction synthesis. Anti-saturation processing prevents the integral term in the PI controller from accumulating continuously when the system output reaches its saturation limit. If anti-saturation processing is not performed when the controller output cannot be further changed due to actuator limitations, the integral term will continue to increase, leading to significant overshoot when the system returns to its normal operating range. Anti-saturation strategies address this problem by limiting the accumulation rate of the integral term or resetting it. Furthermore, a nonlinear mapping is performed on the q-axis integral term to obtain the yaw feedforward signal. Specifically, the q-axis integral term is nonlinearly mapped using the following formula: , in, For the q-axis integral term, Dead zone threshold, For the current gain, This is the saturation limit. This means that when the absolute value of the q-axis integral term is less than the dead zone threshold, the yaw feedforward signal is zero to avoid responding to small disturbances; when its absolute value exceeds the dead zone threshold, the yaw feedforward signal will increase linearly according to the current gain until it reaches the maximum output specified by the saturation limit, thereby effectively limiting the output range of the feedforward signal.

[0025] Specifically, in step S4, based on the azimuth angle, the d-axis pitch compensation command, q-axis pitch compensation command, and collective pitch command are fused to obtain independent pitch commands for the first to third blades. That is, the d-axis and q-axis pitch compensation commands generated by the dual-channel PI controller are transformed from the fixed coordinate system back to the blade rotation coordinate system as independent compensation quantities, and then fused with the collective pitch command to ultimately form individual pitch commands for each blade. This refined independent pitch control can more effectively manage and compensate for blade load imbalances and asymmetries caused by wind shear, tower shadow effects, and yaw errors. In this process, although the dq transformation simplifies the controller design, the actual pitch actuator operates independently for each blade.

[0026] Although the dq transformation simplifies controller design, the actual pitch actuator operates independently for each blade. Therefore, it is necessary to convert the dq-axis compensation commands back to the compensation amounts corresponding to each blade to ensure that each blade obtains the most suitable pitch angle based on its azimuth and stress conditions. Therefore, in the technical solution of this application, firstly, based on the azimuth angle, the inverse Coleman transformation matrix is ​​determined. The inverse Coleman transformation matrix is ​​the inverse of the Coleman transformation matrix, used to convert quantities in the dq fixed coordinate system back to independent blade quantities in the rotating coordinate system. Then, based on the inverse Coleman transformation matrix, the d-axis pitch compensation command and the q-axis pitch compensation command are subjected to an inverse Coleman transformation to obtain the independent variable compensation amounts for the first to third blades. Specifically, the d-axis pitch compensation command and the q-axis pitch compensation command are subjected to an inverse Coleman transformation using the following formula: , in, These are the d-axis pitch compensation commands and the q-axis pitch compensation commands. The compensation amounts for the independent variables of the first to third blades are... It is the inverse Coleman transformation matrix; Furthermore, the independent variable compensation values ​​of the first to third blades are fused with the collective pitch command and rate-limited to obtain the independent pitch commands for the first to third blades. That is, this compensation command is fused with the collective pitch command determined by factors such as wind speed, enabling the optimization of local loads and correction of errors while maintaining overall power output control. This improves the stability of wind turbine operation, reduces fatigue load, and ultimately increases power generation efficiency and extends unit life. Specifically, the independent variable compensation values ​​of the first to third blades are fused with the collective pitch command using the following formula: , in, For collective pitch control commands, The compensation amount is the independent variable for each blade; Following this, rate limiting is required for the merged commands. Rate limiting prevents the pitch actuator from making excessive pitch speed adjustments in a short period, which could lead to excessive mechanical stress, system instability, or shortened actuator life. By setting a maximum permissible pitch rate, smooth pitch action is ensured to remain within a safe range. For example, if the command change exceeds the maximum permissible rate, the actual output is limited to the maximum rate range. The final rate-limited commands are the independent pitch commands for the first to third blades. Specifically, the merged commands are rate-limited using the following formula: , in, It is a constraint function that restricts x to the range [min, max]. It is the pitch control command for the previous time step. It is the maximum permissible pitch rate. It is the sampling time. It is a command that has undergone rate limiting processing, that is, an independent pitch control command for the first to third blades.

[0027] Specifically, S5 involves coordinated control and execution based on the independent pitch commands of the first to third blades, the yaw feedforward signal, and the macroscopic yaw error. It should be understood that wind turbines, in actual operation, are affected by various factors such as wind direction changes, wind shear, and tower shadow effects, leading to asymmetrical blade loads and yaw errors between the nacelle and the wind direction. Without effective compensation, these problems will not only affect power generation efficiency but also accelerate the fatigue wear of mechanical components and may even endanger the safety of the unit. By coordinating and executing the independent pitch commands generated to balance blade loads, the yaw feedforward signal generated to correct yaw in advance, and the macroscopic yaw error reflecting the overall windward deviation of the unit, it is possible to ensure that the wind turbine maintains optimal windward conditions in dynamically changing wind fields, while minimizing blade loads, thereby increasing power generation, reducing operation and maintenance costs, and extending the service life of the wind turbine. This coordinated control and execution reflects the linkage and cooperation between the pitch system and the yaw system, which is key to achieving intelligent wind power control.

[0028] In practice, the first step is to generate and fuse a mixed yaw command from the yaw feedforward signal and the macroscopic yaw error to obtain a mixed target yaw rate command. During this process, the system fuses the yaw feedforward signal and the macroscopic yaw error to generate a comprehensive target yaw rate command. The yaw feedforward signal comes from the q-axis integral term of the dual-channel PI controller, reflecting predictive compensation for the long-term cumulative deviation of the yaw torque; the macroscopic yaw error represents the overall deviation between the current cabin and the actual wind direction. This fusion process aims to combine the rapid response of feedforward control with the corrective capability of feedback control. For example, the macroscopic yaw error can be converted into a basic yaw rate command through a proportional-integral (PI) controller, and then the yaw feedforward signal can be superimposed on this basic command, or fusion can be performed using weighted averaging or other methods. The purpose of fusion is to reduce the hysteresis and overshoot of the yaw system while ensuring accurate wind alignment, thereby improving the smoothness and accuracy of yaw control. Finally, this process outputs a command representing the desired yaw speed. Next, the mixed target yaw rate command undergoes a state machine-based yaw execution decision to obtain the final yaw motor drive command. The mixed target yaw rate command is not directly sent to the yaw motor; instead, it is first input into a state machine-based yaw execution decision module. This state machine is the core of the yaw control system. It analyzes and processes the target yaw rate command based on the wind turbine's current operating state, environmental conditions, safety constraints, and internal logic. For example, the state machine may include states such as: standby, normal windward operation, emergency stop, anti-tangling, and sudden wind direction change. Under different states, the state machine may limit, filter, or modify the target yaw rate command to ensure the safety and rationality of the yaw action. For example, in strong winds or freezing conditions, the yaw speed may be limited; when the wind direction is stable and the error is small, only minor adjustments may be made. Through this state machine-based decision-making process, the system can intelligently determine when and how to execute the yaw operation, ultimately outputting a safe and effective yaw motor drive command that directly controls the yaw motor's speed and direction. Then, the independent pitch commands for the first to third blades and the final yaw motor drive command are executed. That is, the independent pitch commands for the first to third blades, generated in the preceding steps and fused and rate-limited, are sent to the pitch actuators (usually hydraulic cylinders or electric servo motors) of each blade, driving the blades to independently adjust their pitch angle. Simultaneously, the final yaw motor drive command, obtained based on state machine decisions, is sent to the yaw motor driver to control the yaw motor's rotation, thereby adjusting the nacelle's orientation so that the rotor plane is as perpendicular to the wind direction as possible. Through the simultaneous coordinated action of the pitch and yaw systems, the wind turbine can respond in real time to changes in wind conditions and load asymmetry, achieving optimal wind conditions and minimal blade load, thus enabling stable and efficient operation of the entire wind turbine unit.

[0029] In summary, the pitch system yaw error collaborative compensation method according to the embodiments of this application is explained. It obtains the blade bending moment matrix, azimuth angle, and macroscopic yaw error, and uses asymmetric load decoupling and quantization technology to convert the blade load into fixed d-axis tilting moment and q-axis yaw moment components. Subsequently, dual-channel PI control is applied to these moment components to generate d-axis and q-axis pitch compensation commands, and the yaw feedforward signal is cleverly extracted from the q-axis integral term. Finally, based on the azimuth angle, the d-axis and q-axis pitch compensation commands are fused with the collective pitch command to obtain the independent pitch command for each blade. This is then combined with the yaw feedforward signal and macroscopic yaw error for collaborative control and execution, thereby achieving deep collaboration between the pitch and yaw systems. This approach more realistically reflects the rotor's stress state and allows for refined compensation, effectively reducing fatigue loads on the blades and tower base.

[0030] Furthermore, a collaborative compensation system for yaw error in a pitch control system is also provided.

[0031] Figure 3 This is a block diagram of a pitch system yaw error collaborative compensation system according to an embodiment of this application. Figure 3 As shown, the pitch system yaw error collaborative compensation system 300 according to an embodiment of this application includes: a data acquisition module 310, used to acquire the bending moment matrix, azimuth angle, and macroscopic yaw error of the first to third blades; an asymmetric load decoupling and quantization module 320, used to perform asymmetric load decoupling and quantization on the bending moment matrix and azimuth angle of the first to third blades to obtain a fixed d-axis tilt moment component and a fixed q-axis yaw moment component; a dual-channel PI control module 330, used to perform dual-channel PI control on the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to obtain d-axis pitch compensation command, q-axis pitch compensation command, and yaw feedforward signal; a command fusion module 340, used to perform command fusion on the d-axis pitch compensation command, q-axis pitch compensation command, and collective pitch command based on the azimuth angle to obtain independent pitch commands for the first to third blades; and a collaborative control and execution module 350, used to perform collaborative control and execution based on the independent pitch commands of the first to third blades, the yaw feedforward signal, and the macroscopic yaw error.

[0032] As described above, the pitch system yaw error collaborative compensation system 300 according to the embodiments of this application can be implemented in various wireless terminals, such as servers with pitch system yaw error collaborative compensation algorithms. In one possible implementation, the pitch system yaw error collaborative compensation system 300 according to the embodiments of this application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the pitch system yaw error collaborative compensation system 300 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the pitch system yaw error collaborative compensation system 300 can also be one of many hardware modules of the wireless terminal.

[0033] Alternatively, in another example, the pitch system yaw error collaborative compensation system 300 and the wireless terminal can also be separate devices, and the pitch system yaw error collaborative compensation system 300 can be connected to the wireless terminal via wired and / or wireless networks, and transmit interactive information in accordance with an agreed data format.

[0034] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for collaborative compensation of yaw error in a pitch control system, characterized in that, include: Obtain the bending moment matrix, azimuth angle, and macroscopic yaw error of the first to third blades; Asymmetric load decoupling and quantization are performed on the bending moment matrix and azimuth angle of the first to third blades to obtain the fixed d-axis tilting moment component and the fixed q-axis yaw moment component. Dual-channel PI control is performed on the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to obtain the d-axis pitch compensation command, the q-axis pitch compensation command and the yaw feedforward signal; Based on the azimuth angle, the d-axis pitch compensation command, the q-axis pitch compensation command, and the collective pitch command are fused to obtain the independent pitch commands for the first to third blades. Coordinated control and execution are achieved based on the independent pitch control commands of the first to third blades, the yaw feedforward signal, and the macroscopic yaw error.

2. The method for coordinated compensation of yaw error in a pitch control system according to claim 1, characterized in that, Obtain the bending moment matrix, azimuth angle, and macroscopic yaw error of the first to third blades, including: The difference between the wind direction angle and the cabin yaw angle is calculated as the original error; The macroscopic yaw error is obtained by normalizing the original error using the following formula: ， in, This is the original error. For finding the remainder function.

3. The method for coordinated compensation of yaw error in a pitch control system according to claim 1, characterized in that, Asymmetric load decoupling and quantization are performed on the bending moment matrix and azimuth angle of the first to third blades to obtain the fixed d-axis tilting moment component and the fixed q-axis yaw moment component, including: Determine the Coleman transformation matrix based on the azimuth angle; Based on the Coleman transformation matrix, the bending moment matrices of the first to third blades are subjected to load projection from a rotating coordinate system to a fixed coordinate system to obtain the unscaled dq axis moment component vectors. Torque components are extracted and power invariant scaling is performed on the unscaled dq-axis torque component vector to obtain the fixed d-axis tilting torque component and the fixed q-axis yaw torque component.

4. The method for coordinated compensation of yaw error in a pitch control system according to claim 3, characterized in that, Based on the Coleman transformation matrix, the bending moment matrices of the first to third blades are subjected to load projection from a rotated coordinate system to a fixed coordinate system to obtain the unscaled dq-axis moment component vectors. This includes: performing load projection from a rotated coordinate system to a fixed coordinate system on the bending moment matrices of the first to third blades using the following formula: ， in, Here is the Coleman transformation matrix. Here is the bending moment matrix for the first to third blades.

5. The method for coordinated compensation of yaw error in a pitch control system according to claim 1, characterized in that, Dual-channel PI control is applied to the fixed d-axis tilt moment component and the fixed q-axis yaw moment component to obtain the d-axis pitch compensation command, the q-axis pitch compensation command, and the yaw feedforward signal, including: The fixed d-axis tilting moment component and the fixed q-axis yaw moment component are used as control errors and input into the PI controller to obtain the d-axis proportional term, d-axis integral term, q-axis proportional term, and q-axis integral term; The d-axis proportional term, d-axis integral term, q-axis proportional term, and q-axis integral term are processed to synthesize commands and output with anti-saturation processing to obtain the d-axis pitch compensation command and the q-axis pitch compensation command. The yaw feedforward signal is obtained by performing a nonlinear mapping on the q-axis integral term.

6. The method for coordinated compensation of yaw error in a pitch control system according to claim 5, characterized in that, To obtain the yaw feedforward signal, the q-axis integral term is nonlinearly mapped using the following formula: ， in, For the q-axis integral term, Dead zone threshold, For the current gain, This is the saturation limit.

7. The method for coordinated compensation of yaw error in a pitch control system according to claim 1, characterized in that, Based on the azimuth angle, the d-axis pitch compensation command, the q-axis pitch compensation command, and the collective pitch command are fused to obtain independent pitch commands for the first to third blades, including: Determine the inverse Coleman transformation matrix based on the azimuth angle; Based on the inverse Coleman transformation matrix, the d-axis pitch compensation command and the q-axis pitch compensation command are subjected to inverse Coleman transformation to obtain the independent variable compensation amounts of the first to third blades. The independent variable compensation values ​​of the first to third blades are respectively fused with the collective pitch command and rate limited to obtain the independent pitch commands of the first to third blades.

8. The method for coordinated compensation of yaw error in a pitch control system according to claim 1, characterized in that, Based on the inverse Coleman transformation matrix, an inverse Coleman transformation is performed on the d-axis pitch compensation command and the q-axis pitch compensation command to obtain the independent variable compensation amounts of the first to third blades. This includes performing an inverse Coleman transformation on the d-axis pitch compensation command and the q-axis pitch compensation command using the following formula: ， in, These are the d-axis pitch compensation commands and the q-axis pitch compensation commands. The compensation amounts for the independent variables of the first to third blades are... It is the inverse Coleman transformation matrix.

9. The method for coordinated compensation of yaw error in a pitch control system according to claim 1, characterized in that, Coordinated control and execution based on independent pitch commands from the first to third blades, yaw feedforward signals, and macroscopic yaw errors, including: The yaw feedforward signal and macroscopic yaw error are used to generate and fuse a mixed yaw command to obtain a mixed target yaw rate command. A state machine-based yaw execution decision is made on the mixed target yaw rate command to obtain the final yaw motor drive command. Execute the independent pitch control commands for the first to third blades and the final yaw motor drive command.

Images