An adaptive lead-pitch control method, system, device and medium for a wind turbine

Abstract

The application relates to the technical field of wind power generation control, in particular to a wind turbine adaptive advance variable pitch control method, system, device and medium, the method comprising: acquiring a characteristic parameter representing a wind speed change trend in front of a wind turbine; identifying an actual response delay of a variable pitch actuator based on a time sequence difference between a variable pitch instruction issuing time and an actual pitch angle change time; dynamically calculating a current advance time based on the characteristic parameter and the actual response delay; performing time delay processing on a front wind speed signal based on the current advance time to generate a predicted wind speed and generate a feedforward variable pitch instruction; and feeding back and correcting calculation parameters of the current advance time based on a wind turbine operation state index after a variable pitch action. The application realizes adaptive closed-loop optimization of the advance time, effectively suppresses speed overshoot and reduces wind turbine fatigue load.

Description

Technical Field

[0001] This disclosure belongs to the field of wind power generation control technology, and in particular relates to an adaptive advance pitch control method, system, equipment and medium for wind turbines. Background Technology

[0002] In the operation and control of wind turbine generators, the pitch system is a key actuator for regulating power output and ensuring generator safety. Due to the large inertia and slow dynamic characteristics of wind turbine generators, while wind speed, especially turbulent wind, changes rapidly and randomly, traditional feedback pitch control based on generator speed deviation often suffers from lag, which can easily lead to speed overshoot and load fluctuations.

[0003] To address this issue, existing technologies have incorporated lidar wind measurement to achieve advance pitch control. However, most current advance pitch strategies employ fixed lead times or simple wind speed differential feedforward, failing to adequately consider the complexity of actual operating conditions. On one hand, the rate of change of wind speed and turbulence intensity vary in real time, making it difficult to match wind conditions of varying severity with a fixed lead time. On the other hand, the response delay of the pitch actuator dynamically changes due to factors such as hydraulic oil temperature and mechanical wear, leading to a mismatch between the preset lead time and the actual execution capability. This open-loop or semi-open-loop control method is prone to problems such as "under-lead" causing large fluctuations in rotational speed, or "over-lead" causing frequent pitch movements and increased component fatigue loads, lacking an effective adaptive closed-loop optimization mechanism. Summary of the Invention

[0004] To address the aforementioned issues, this disclosure provides an adaptive lead pitch control method, system, equipment, and medium for wind turbines. It employs an adaptive closed-loop control scheme that can identify response delays in real time, dynamically adjust the lead time according to wind conditions, and combine the feedback of the unit's operating status to correct parameters. This effectively solves the problem of fixed lead time in existing technologies, which cannot adapt to changes in actual operating conditions, thereby suppressing speed overshoot and reducing fatigue load on the unit.

[0005] Firstly, this disclosure provides an adaptive advance pitch control method for wind turbine generators. The method includes, Obtain characteristic parameters that characterize the trend of wind speed change in front of the wind turbine; Based on the timing difference between the pitch command issuance time and the actual pitch angle change time, the actual response delay of the pitch actuator is identified. Calculate the current lead time based on the characteristic parameters and the actual response delay; The wind speed signal ahead is delayed based on the current lead time to generate a predicted wind speed, and a feedforward pitch command is generated based on the predicted wind speed. Based on the unit's operating status indicators after pitch control, the calculation parameters for the current lead time are corrected.

[0006] Furthermore, Obtain characteristic parameters that characterize the trend of wind speed variation in front of the wind turbine, specifically including: Wind speed data is collected in real time at multiple preset distance gates using forward-looking lidar; Based on the wind speed data at multiple distance gates, the equivalent wind speed on the impeller surface is calculated; The rate of change of wind speed is calculated based on the ratio of the difference between the equivalent wind speeds on the impeller surface at adjacent sampling times to the sampling time interval. Calculate the standard deviation of wind speed within a preset time window; The wind speed change rate and the wind speed standard deviation are used as the characteristic parameters.

[0007] Furthermore, Based on the timing difference between the pitch command issuance time and the actual pitch angle change time, the actual response delay of the pitch actuator is identified, specifically including: After the pitch command is issued, the actual pitch angle is continuously monitored. The moment when the actual pitch angle first deviates from the preset dead zone is determined as the moment when the actual pitch angle changes. Calculate the time difference between the moment the pitch command is issued and the moment the actual pitch angle changes, and filter the time differences obtained multiple times to obtain an estimated response delay under the current operating condition, which is then used as the actual response delay.

[0008] Furthermore, Based on the characteristic parameters and the actual response delay, the current lead time is calculated, specifically including: Obtain the baseline lead time, wind speed change rate correction term, turbulence intensity correction term, and response delay correction term; The initial lead time is obtained by superimposing the baseline lead time, the wind speed change rate correction term, the turbulence intensity correction term, and the response delay correction term. Based on the preset upper and lower limits of the lead time, the initial lead time is constrained to obtain the current lead time; The wind speed change rate correction term is determined based on the wind speed change rate in the characteristic parameters, the turbulence intensity correction term is determined based on the wind speed standard deviation in the characteristic parameters, and the response delay correction term is determined based on the deviation between the actual response delay and the nominal response delay.

[0009] Furthermore, The wind speed change rate correction term is proportional to the product of the absolute value of the wind speed change rate and the first proportionality coefficient; The turbulence intensity correction term is proportional to the product of the wind speed standard deviation and the second proportionality coefficient; The response delay correction term is proportional to the product of the deviation of the actual response delay from the nominal response delay and the third proportionality coefficient; The reference lead time is obtained by calibrating the blade length of the wind turbine and the aerodynamic response time at the rated wind speed.

[0010] Furthermore, Based on the current lead time, the forward wind speed signal is delayed to generate a predicted wind speed, and a feedforward pitch command is generated based on the predicted wind speed, specifically including: The wind speed signal at the farthest point of the forward lidar is selected as the forward wind speed signal. The forward wind speed signal is delayed by the current lead time and used as the predicted value of the future wind speed on the impeller surface to obtain the predicted wind speed. Based on the predicted wind speed, current generator speed and current pitch angle, the required pitch command is calculated and superimposed on the output of the original feedback pitch controller. When the lidar data used to acquire the feature parameters fails, it automatically switches to pure feedback pitch mode; When the actual response delay exceeds a preset reasonable range, the maximum lead time is limited and a maintenance alarm is triggered.

[0011] Furthermore, Based on the unit's operating status indicators after pitch control, the calculation parameters for the current lead time are corrected through feedback, specifically including: The maximum overshoot of generator speed, the time to recover to a stable speed, and the fatigue cost of pitch control after pitch adjustment are collected as the operating status indicators of the unit. When the maximum overshoot of the generator speed is greater than the first preset threshold, or the time for the speed to recover to a stable state is greater than the second preset threshold, it is determined to be an under-lead state. When the pitch control frequency is higher than the historical average and the maximum overshoot of the generator speed is less than the third preset threshold, it is determined to be an over-leading state. Based on the under-leading state or the over-leading state, the first proportional coefficient, the second proportional coefficient, the third proportional coefficient, and / or the reference lead time are adjusted in a closed loop.

[0012] Secondly, based on the same inventive concept, this disclosure provides an adaptive advance pitch control system for a wind turbine, comprising: The wind condition sensing module is used to acquire characteristic parameters that characterize the trend of wind speed changes in front of the wind turbine. The delay identification module is used to identify the actual response delay of the pitch actuator based on the timing difference between the pitch command issuance time and the actual pitch angle change time. The time calculation module is used to calculate the current lead time based on the characteristic parameters and the actual response delay; The instruction generation module is used to perform delay processing on the forward wind speed signal based on the current lead time to generate a predicted wind speed, and generate a feedforward pitch instruction based on the predicted wind speed. The feedback correction module is used to correct the calculation parameters of the current lead time based on the unit's operating status indicators after the pitch change operation.

[0013] Thirdly, this disclosure also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When the processor executes the program stored in the memory, it implements the steps of the adaptive advance pitch control method for any of the wind turbine units described above.

[0014] Fourthly, this disclosure also provides a computer storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the adaptive advance pitch control methods for wind turbine generators as described above.

[0015] Compared with the prior art, this disclosure has the following advantages: By acquiring characteristic parameters representing wind speed change trends and identifying the actual response delay of the pitch actuator online, a dynamic mapping relationship between the severity of external wind conditions and the response capability of the internal mechanism was established, solving the problem that fixed lead time cannot adapt to time-varying operating conditions. Based on this, the current lead time is calculated by superimposing multiple maintenance positive items, enabling the control timing to be adjusted in real time according to the wind speed change rate, turbulence intensity, and the aging degree of the mechanism. Based on the feedback correction mechanism of the unit's operating status indicators after pitch control, the system can automatically identify "under-lead" or "over-lead" states and perform closed-loop optimization of the calculation parameters according to the actual effects such as speed overshoot, settling time, and fatigue costs. This complete closed-loop logic of "perception-decision-execution-evaluation-correction" not only effectively suppresses speed overshoot under gusts or strong turbulence conditions but also avoids frequent pitch control actions caused by excessive lead control, thereby significantly reducing the fatigue load on key components and extending the unit's service life while improving power generation quality.

[0016] Other features and advantages of this disclosure will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the disclosure. The objects and other advantages of this disclosure may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating an adaptive advance pitch control method for a wind turbine according to an embodiment of the present disclosure is shown. Figure 2 A structural block diagram of an adaptive advance pitch control system for a wind turbine according to an embodiment of the present disclosure is shown. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0020] Figure 1 A flowchart illustrating an adaptive advance pitch control method for a wind turbine according to an embodiment of the present disclosure is shown, as follows: Figure 1 As shown in the present disclosure, the adaptive advance pitch control method for wind turbines includes, S1, obtain the characteristic parameters that characterize the trend of wind speed change in front of the wind turbine; The characteristic parameters of this disclosure can include indicators that quantify the severity and direction of wind conditions, such as wind speed change rate, turbulence intensity, wind shear coefficient, or combinations thereof. For example, wind field data in the foreground space can be collected using lidar or multi-point anemometers, and the trend quantity characterizing the rate of wind speed change over time can be extracted after signal processing. This embodiment extracts information that reflects the changing trend of wind energy input over a future period, rather than simply obtaining the instantaneous wind speed value at the current moment. By obtaining such trend characteristic parameters, the control system can detect upcoming aerodynamic excitation changes in advance, providing a forward-looking input basis for subsequent proactive control, thereby overcoming the hysteresis defect of relying solely on current wind speed for feedback adjustment.

[0021] In this embodiment of the disclosure, step S1 specifically includes: S11 uses forward-facing lidar to collect wind speed data in real time at multiple preset distance gates; S12, Calculate the equivalent wind speed on the impeller surface based on the wind speed data at multiple distance gates; S13, Calculate the rate of change of wind speed based on the ratio of the difference of the equivalent wind speed on the impeller surface at adjacent sampling times to the sampling time interval; S14, calculate the standard deviation of wind speed within a preset time window; S15, the wind speed change rate and the wind speed standard deviation are used as the characteristic parameters.

[0022] In this embodiment, a forward-facing lidar is installed on the top of the wind turbine nacelle. Multiple "distance gates" (e.g., 30 m, 60 m, 90 m, 120 m) are positioned at different distances in front of the rotor. Each distance gate corresponds to a specific spatial location, and the lidar measures the wind speed at that location in real time. By measuring through multiple distance gates, the spatial distribution of wind speed along the incoming flow direction and its variation over time can be obtained, providing a data basis for predicting the time it takes for wind speed to reach the rotor surface.

[0023] In this embodiment, the wind speed measured at different distance gates may differ due to factors such as turbulence and wind shear. To reflect the overall wind speed that will act on the impeller's plane of rotation, the wind speed values ​​from multiple distance gates need to be weighted or averaged to obtain an "equivalent wind speed." The equivalent wind speed is typically calculated using spatial averaging or weighted calculation based on the impeller's aerodynamic model, and is used to represent the overall wind speed level felt by the impeller surface at the current moment.

[0024] In this embodiment of the disclosure, the rate of change of wind speed reflects the degree of drastic change in wind speed over time. The calculation method is as follows: In the formula, V eq The impeller surface equivalent wind speed, The sampling period.

[0025] In this embodiment of the disclosure, the standard deviation of the equivalent wind speed sequence is calculated within a short time window (e.g., 10 seconds). The standard deviation is a statistical indicator that measures the degree of wind speed fluctuation and can characterize the intensity of turbulence. The larger the standard deviation, the more severe the wind speed fluctuation, the more complex the wind conditions, and the more necessary the intervention of proactive control.

[0026] S2, based on the timing difference between the pitch command issuance time and the actual pitch angle change time, identify the actual response delay of the pitch actuator; Specifically, after receiving a control command, the actual action of the pitch actuator (such as a hydraulic system or electric pitch drive) often lags behind the moment the command is issued due to physical factors such as mechanical transmission backlash, hydraulic compressibility, or motor starting characteristics. This application's embodiment quantifies the current response delay by monitoring and recording the time difference between two key moments in real time. This online identification method based on measured data can accurately reflect the dynamic performance of the pitch system under specific operating conditions (such as low temperature and high viscosity, component wear and aging), avoiding errors caused by using factory nominal values ​​or fixed empirical values, and ensuring the accuracy of the lead time calculation benchmark.

[0027] In this embodiment of the disclosure, step S2 specifically includes: S21, after the pitch command is issued, continuously monitor the actual pitch angle; S22, the moment when the actual pitch angle first deviates from the preset dead zone is determined as the moment when the actual pitch angle changes; S23, calculate the time difference between the moment the pitch command is issued and the moment the actual pitch angle changes, and filter the time difference obtained multiple times to obtain an estimated response delay under the current operating condition, which is used as the actual response delay.

[0028] In this embodiment of the disclosure, the angle range of the preset dead zone is 0.1 degrees to 0.3 degrees.

[0029] In this embodiment of the disclosure, the filtering process for the time difference acquired multiple times consecutively includes: A sliding window mid-value filtering algorithm is used to remove outlier sampled data; When the estimated response delay value continuously exceeds the preset delay threshold, a pitch system health warning is triggered.

[0030] In this embodiment of the disclosure, abnormal sampling data refers to data points whose deviation from the median within the current sliding window exceeds a preset multiple, such as an absolute deviation from the median within the window greater than 1.5 times the median of absolute deviations within the window; The preset delay threshold is 500ms to 800ms (which can be calibrated according to the specific machine model and hydraulic system characteristics, for example, 600ms).

[0031] In this embodiment of the disclosure, by introducing a preset dead zone as a threshold for determining effective action, after the pitch command is issued, the system continuously collects the feedback signal from the pitch angle sensor, but does not immediately regard small numerical fluctuations as the start of the response. Only when the actual pitch angle change exceeds the boundary of the preset dead zone is the moment marked as the actual pitch angle change moment. This can effectively filter out these non-functional disturbances and ensure that the captured delay truly reflects the dynamic process of the actuator overcoming static friction and generating effective aerodynamic torque.

[0032] S3, Calculate the current lead time based on the characteristic parameters and the actual response delay; In this embodiment, the current lead time is not determined by a preset fixed value, but rather by real-time adaptation based on the intensity of external wind conditions and the responsiveness of internal actuators. When characteristic parameters show a drastic trend in wind speed changes, it means that earlier intervention is needed to suppress rotational speed fluctuations; when the actual response delay increases, it means that the actuator's action slows down, and similarly, instructions need to be issued earlier to compensate for this additional waiting time. This embodiment establishes a dynamic mapping relationship for lead time by using characteristic parameters representing external environmental demands and the actual response delay representing internal execution capabilities as dual input variables. This dual-driven dynamic calculation mechanism allows the control strategy to adaptively adjust to changes in wind randomness and equipment health status, effectively preventing under-lead or over-lead problems caused by lead time mismatch.

[0033] In this embodiment of the disclosure, step S3 specifically includes: S31, obtain the baseline lead time, wind speed change rate correction term, turbulence intensity correction term, and response delay correction term; S32, the reference lead time, the wind speed change rate correction term, the turbulence intensity correction term, and the response delay correction term are superimposed to obtain the initial lead time; S33, Based on the preset upper and lower limits of the lead time, the initial lead time is constrained to obtain the current lead time; The wind speed change rate correction term is determined based on the wind speed change rate in the characteristic parameters, the turbulence intensity correction term is determined based on the wind speed standard deviation in the characteristic parameters, and the response delay correction term is determined based on the deviation between the actual response delay and the nominal response delay.

[0034] In this embodiment of the disclosure, the wind speed change rate correction term is proportional to the product of the absolute value of the wind speed change rate and the first proportionality coefficient; The turbulence intensity correction term is proportional to the product of the wind speed standard deviation and the second proportionality coefficient; The response delay correction term is proportional to the product of the deviation of the actual response delay from the nominal response delay and the third proportionality coefficient; The reference lead time is obtained by calibrating the blade length of the wind turbine and the aerodynamic response time at the rated wind speed.

[0035] In this embodiment of the disclosure, the preset lower limit of the lead time is 0.5 seconds, and the preset upper limit of the lead time is 4 seconds.

[0036] In this embodiment, the absolute value of the wind speed change rate directly characterizes the degree of drastic change in input wind energy per unit time, i.e., the rate of change of aerodynamic excitation intensity. When the wind speed rises or falls sharply, the rate of change of aerodynamic torque on the wind turbine increases. If the original lead time is maintained, the pitch control action may not be able to offset this rapidly changing torque disturbance in time, resulting in large fluctuations in rotational speed. Therefore, by introducing a correction term proportional to the absolute value of the wind speed change rate, the system can automatically extend the lead time under gusts or strong shear wind conditions, initiating the pitch control action earlier to match the stronger aerodynamic excitation change requirements. The first proportional coefficient is used to adjust the sensitivity of this correction term and can be tuned according to the unit's rotational inertia and the maximum pitch rate of the pitch control actuator. The first proportional coefficient can be taken as 0.3~0.5s. 2 / m.

[0037] In this embodiment, the wind speed standard deviation is a core statistical indicator for measuring turbulence intensity. Its physical essence reflects the distribution width and random fluctuation amplitude of wind field energy in the frequency domain. High turbulence means that the wind speed signal contains more high-frequency random components. While these components may not necessarily manifest as macroscopic wind speed trend changes, they continuously excite vibration modes in the tower and drivetrain. In this case, appropriately increasing the lead time is equivalent to introducing a low-pass filter effect, making the pitch command smoother and more forward-looking, avoiding ineffective and frequent pitch movements caused by chasing high-frequency noise. The setting of the second proportional coefficient needs to balance vibration suppression and power generation efficiency, preventing excessive smoothing from causing a delay in capturing effective wind energy. The second proportional coefficient can be set to 0.8~1.2 s / m.

[0038] In this embodiment, the nominal response delay typically refers to the standard response time of the pitch system under factory testing or ideal operating conditions. However, in actual operation, factors such as changes in hydraulic oil temperature, valve body wear, and increased gear clearance can all cause the actual response delay to drift. When the actual response delay is greater than the nominal value, it indicates that the actuator is moving slower. If no compensation is made, the originally calculated lead time will become invalid due to this additional mechanical lag, resulting in the actual control timing being later than expected. By introducing this correction term, the system can convert the identified mechanical performance degradation into a time compensation amount in real time, ensuring that even under conditions of actuator aging or harsh environments, the actual effective time of the lead control can still accurately align with the wind condition change node. The third proportional coefficient is typically set to a value close to 1 to ensure the consistency of the compensation amount and the delay deviation on the time scale. The third proportional coefficient can be 0.9 to 1.1.

[0039] It should be understood that although this embodiment employs a computational structure of linear superposition of four correction terms, this is not the only implementation. In other embodiments, nonlinear relationships such as weighted fusion, lookup table mapping, or polynomial fitting can also be used between the correction terms to determine the current lead time. For example, a two-dimensional lookup table with wind speed change rate and response delay as input can be pre-established through simulation or actual measurement to directly map the output of the optimal lead time, thereby avoiding the computational burden caused by complex formula calculations; or a neural network model can be used to learn the coupling relationship between various factors. As long as the purpose of dynamically adjusting the lead time based on feature parameters and actual response delay can be achieved, the specific mathematical expression should not constitute a limitation of this application.

[0040] S4, based on the current lead time, perform delay processing on the forward wind speed signal to generate a predicted wind speed, and generate a feedforward pitch command based on the predicted wind speed. In this embodiment, after determining the optimal lead time, the system shifts or delays the wind speed signal measured ahead on the time axis to reconstruct the expected wind speed distribution on the rotor surface at future moments, i.e., the predicted wind speed. Subsequently, based on this predicted wind speed and the current operating status of the unit (such as speed and pitch angle), the required pitch angle or pitch rate is calculated in advance using an aerodynamic model or lookup table method, forming a feedforward pitch command. This allows the pitch system to begin operating before gusts actually reach the rotor, transforming traditional passive feedback regulation into active feedforward suppression, significantly improving the unit's ability to cope with rapidly changing wind conditions.

[0041] In this embodiment of the disclosure, step S4 specifically includes: S41, Select the wind speed signal at the farthest door of the forward lidar as the forward wind speed signal; S42, the forward wind speed signal is delayed by the current lead time and used as the predicted value of the future wind speed on the impeller surface to obtain the predicted wind speed; S43, based on the predicted wind speed, current generator speed and current pitch angle, calculate the required pitch command and superimpose the required pitch command onto the output of the original feedback pitch controller; S44, When the lidar data used to acquire the feature parameters fails, it automatically switches to pure feedback pitch mode; S45, when the actual response delay exceeds a preset reasonable range, limit the maximum lead time and trigger a maintenance alarm.

[0042] In this embodiment, among the multiple distance gates (e.g., 30 m, 60 m, 90 m, 120 m) set by the forward-facing lidar, the wind speed signal corresponding to the farthest distance gate (e.g., 120 m) contains the wind condition information at the forefront. This location has the largest physical distance to the impeller surface, resulting in the longest wind propagation time, thus providing sufficient "predictive time" for forward control. Selecting the farthest distance gate as the original wind speed signal fully utilizes the lidar's detection range, maximizing the forward-looking nature of the control.

[0043] In this embodiment of the disclosure, the predicted wind speed, current generator speed, and current pitch angle are incorporated into the control algorithm to calculate the required pitch command. Common methods include: MPPT lookup table method: The optimal pitch angle command is obtained by directly looking up a table based on the predicted wind speed; Model Predictive Control (MPC): Calculates the optimal pitch sequence through short-time domain optimization; The feedforward pitch command is added to the output of the existing feedback pitch controller (such as a PID controller) to form the final command, which is the required pitch command.

[0044] In this embodiment, the LiDAR may experience a low signal-to-noise ratio due to rain, fog, snow, icing, or optical window contamination, rendering the data invalid. In such cases, the forward wind speed sequence cannot be obtained, and characteristic parameters (wind speed change rate, standard deviation) are unavailable. The system should automatically switch to the traditional pure feedback pitch mode (i.e., using only the generator speed deviation as input and no longer generating feedforward pitch commands) to ensure safe and continuous operation of the unit. The switching process must be smooth (e.g., linear weighted transition) to avoid abrupt command changes.

[0045] In the implementation of this disclosure, a reasonable preset range is 100ms to 1000ms; If the response exceeds the preset reasonable range, it indicates that the pitch actuator is severely lagging (e.g., due to hydraulic system failure or bearing jamming). In this case, the maximum value of the lead time should be actively limited (e.g., reduced from 4 seconds to 2 seconds) to avoid control instability caused by over-prediction. Simultaneously, a maintenance alarm should be triggered to prompt maintenance personnel to inspect the pitch system.

[0046] S5, based on the unit operating status indicators after the pitch control action, the calculation parameters of the current lead time are fed back and corrected.

[0047] In this embodiment, after the pitch control action is completed, operating status indicators such as generator speed overshoot, settling time, or load fluctuation are collected as the basis for performance evaluation. Crucially, the feedback correction targets the underlying parameters used to calculate the lead time (such as the proportional gain and reference time), rather than directly modifying the output pitch command. This ensures both the continuity and safety of each control action and endows the system with the adaptive capability to continuously approach optimal control performance during long-term operation, thereby achieving continuous iterative improvement in control performance.

[0048] In this embodiment of the disclosure, step S5 specifically includes: S51, collect the maximum overshoot of generator speed, the time to recover and stabilize speed, and the fatigue cost of pitch control after pitch control as the operating status indicators of the unit. S52, when the maximum overshoot of the generator speed is greater than the first preset threshold, or the time for the speed to recover to a stable state is greater than the second preset threshold, it is determined to be an under-leading state; S53, when the pitch control frequency is higher than the historical average and the maximum overshoot of the generator speed is less than the third preset threshold, it is determined to be an over-leading state. S54, based on the under-leading state or the over-leading state, perform closed-loop adjustment on the first proportional coefficient, the second proportional coefficient, the third proportional coefficient and / or the reference lead time.

[0049] In this embodiment, the maximum overshoot of the generator speed directly reflects the ability of the pitch control to suppress aerodynamic torque disturbances. If this value is too large, it indicates that the pitch intervention is too late or insufficient, failing to effectively offset the energy impact caused by sudden wind speed changes. The speed recovery time characterizes the dynamic quality of the system returning to steady state from the disturbance; an excessively long time indicates insufficient damping or control phase lag. It is particularly important to note that the fatigue cost of the pitch control action is a key indicator for evaluating the economy of the control strategy. In this embodiment, it is quantified as the cumulative pitch travel per unit time or the integral of the work done by the pitch motor. For example, the cumulative travel can be obtained by integrating the absolute value of the pitch rate within a preset time window, or the work done can be obtained by integrating the product of the pitch motor current and the speed over time. This prevents the control system from excessively frequently driving the pitch mechanism in pursuit of ultimate speed stability, thereby seeking an optimal balance between power generation quality and component lifespan.

[0050] In this embodiment, the boundary for determining the under-lead state is set when the maximum overshoot of the generator speed exceeds a first preset threshold, or the time for the speed to recover to a stable state exceeds a second preset threshold. This typically corresponds to scenarios where wind speed changes drastically but the lead time is set too short, causing the pitch control to lag behind wind changes and failing to unload aerodynamic loads in time. Conversely, the boundary for determining the over-lead state is set when the pitch control frequency is higher than the historical average and the maximum overshoot of the generator speed is less than a third preset threshold. This often occurs when wind conditions are relatively stable but the lead time is set too long, causing the system to over-predict small wind speed fluctuations. Although the speed fluctuations are suppressed to a very low level, unnecessary mechanical wear is incurred. This dual-boundary determination mechanism avoids the one-sidedness of a single indicator evaluation and ensures the accuracy of the adaptive adjustment direction.

[0051] In this embodiment of the disclosure, the first preset threshold can be set to 2% of the rated speed (approximately 36 rpm for a rated speed of 1800 rpm). The second preset threshold can be set to 8 seconds; The third preset threshold can be set to 0.5% of the rated speed.

[0052] In this embodiment, when under-lead is detected, increasing the proportional gain allows the system to calculate a longer lead time under the same wind conditions, thereby accelerating the pitch intervention rhythm; conversely, decreasing sensitivity reduces ineffective actions. To prevent frequent jumps in the proportional gain due to sensor noise or transient disturbances, this embodiment introduces an integral correction strategy with a dead zone during the adjustment process. Specifically, the integrator only begins to accumulate and output the correction amount after the error of the state index deviating from the target value continuously exceeds the dead zone range and remains so for a certain period of time. This design is equivalent to adding a low-pass filter to the parameter update channel, which not only ensures the ability to track changes in real operating conditions but also effectively filters out parameter jitter caused by high-frequency interference, improving the robustness of the control system.

[0053] In this embodiment, the self-tuning principle for the reference lead time focuses on the performance evolution throughout the entire lifecycle of the unit. As blade surface roughness increases, gearbox wears, or the hydraulic system ages, the aerodynamic response and mechanical actuation characteristics of the unit will slowly drift. By statistically analyzing the average lead time, average speed fluctuations, and cumulative pitch stroke over long periods (such as daily or weekly), trend terms reflecting changes in the unit's inherent characteristics can be extracted, and the reference lead time can be fine-tuned accordingly. For example, if long-term statistics show that the average lead time required under the same wind conditions is trending upwards, the reference value is automatically adjusted upwards to compensate for the overall system response lag. This slow self-tuning mechanism is decoupled from and complements the scaling factor adjustment of the fast layer. The former is responsible for compensating for gradual hardware aging, while the latter is responsible for responding to instantaneous environmental changes, together forming an adaptive optimization closed loop that combines agility and stability.

[0054] It should be understood that although this embodiment describes adjustment logic based on a specific threshold and step size, in other embodiments, more advanced adaptive algorithms such as fuzzy control, reinforcement learning, or model predictive control can be used to achieve closed-loop optimization of parameters. For example, a neural network can be used to establish a nonlinear mapping relationship between state indicators and optimal parameters, replacing the fixed linear step size adjustment rule. Furthermore, the quantification method of fatigue cost is not limited to travel accumulation or work integration; it can also be based on equivalent fatigue load spectra such as rainflow counting. As long as the goal of dynamically correcting the lead time calculation parameters based on operational feedback can be achieved, the specific algorithm form and indicator definition should not constitute a limitation of this invention.

[0055] Based on the above method, this disclosure also provides an adaptive advance pitch control system for wind turbines corresponding to the above method. Figure 2 A structural block diagram of an adaptive advance pitch control system for a wind turbine according to an embodiment of the present disclosure is shown. See also: Figure 2 As shown, it includes: Wind condition sensing module 10 is used to acquire characteristic parameters that characterize the trend of wind speed change in front of the wind turbine. The delay identification module 20 is used to identify the actual response delay of the pitch actuator based on the timing difference between the pitch command issuance time and the actual pitch angle change time. The time calculation module 30 is used to calculate the current lead time based on the feature parameters and the actual response delay; The instruction generation module 40 is used to perform delay processing on the forward wind speed signal based on the current lead time to generate a predicted wind speed, and generate a feedforward pitch instruction based on the predicted wind speed. The feedback correction module 50 is used to correct the calculation parameters of the current lead time based on the unit operating status indicators after the pitch change action.

[0056] Based on the same inventive concept as the above disclosure, this disclosure also provides an electronic device. The electronic device of this disclosure includes at least one processor and at least one memory electrically connected to the processor. The memory is electrically connected to the processor, wherein the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method described above.

[0057] It should be noted that the electrical connection between the above-mentioned units does not necessarily mean the connection between lines. The indirect connection method can be applied to the embodiments of this disclosure as long as it achieves the purpose of this disclosure.

[0058] Based on the same inventive concept, this disclosure also provides a computer storage medium storing a computer program, which, when executed by a processor, implements the steps of the above method.

[0059] Although the present disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. An adaptive advance pitch control method for wind turbine generators, characterized in that, The method includes, Obtain characteristic parameters that characterize the trend of wind speed change in front of the wind turbine; Based on the timing difference between the pitch command issuance time and the actual pitch angle change time, the actual response delay of the pitch actuator is identified. Calculate the current lead time based on the characteristic parameters and the actual response delay; The wind speed signal ahead is delayed based on the current lead time to generate a predicted wind speed, and a feedforward pitch command is generated based on the predicted wind speed. Based on the unit's operating status indicators after pitch control, the calculation parameters for the current lead time are corrected.

2. The method according to claim 1, characterized in that, Obtain characteristic parameters that characterize the trend of wind speed variation in front of the wind turbine, specifically including: Wind speed data is collected in real time at multiple preset distance gates using forward-looking lidar; Based on the wind speed data at multiple distance gates, the equivalent wind speed on the impeller surface is calculated; The rate of change of wind speed is calculated based on the ratio of the difference between the equivalent wind speeds on the impeller surface at adjacent sampling times to the sampling time interval. Calculate the standard deviation of wind speed within a preset time window; The wind speed change rate and the wind speed standard deviation are used as the characteristic parameters.

3. The method according to claim 1, characterized in that, Based on the timing difference between the pitch command issuance time and the actual pitch angle change time, the actual response delay of the pitch actuator is identified, specifically including: After the pitch command is issued, the actual pitch angle is continuously monitored. The moment when the actual pitch angle first deviates from the preset dead zone is determined as the moment when the actual pitch angle changes. Calculate the time difference between the moment the pitch command is issued and the moment the actual pitch angle changes, and filter the time differences obtained multiple times to obtain an estimated response delay under the current operating condition, which is then used as the actual response delay.

4. The method according to claim 1, characterized in that, Based on the characteristic parameters and the actual response delay, the current lead time is calculated, specifically including: Obtain the baseline lead time, wind speed change rate correction term, turbulence intensity correction term, and response delay correction term; The initial lead time is obtained by superimposing the baseline lead time, the wind speed change rate correction term, the turbulence intensity correction term, and the response delay correction term. Based on the preset upper and lower limits of the lead time, the initial lead time is constrained to obtain the current lead time; The wind speed change rate correction term is determined based on the wind speed change rate in the characteristic parameters, the turbulence intensity correction term is determined based on the wind speed standard deviation in the characteristic parameters, and the response delay correction term is determined based on the deviation between the actual response delay and the nominal response delay.

5. The method according to claim 4, characterized in that, The wind speed change rate correction term is proportional to the product of the absolute value of the wind speed change rate and the first proportionality coefficient; The turbulence intensity correction term is proportional to the product of the wind speed standard deviation and the second proportionality coefficient; The response delay correction term is proportional to the product of the deviation of the actual response delay from the nominal response delay and the third proportionality coefficient; The reference lead time is obtained based on the blade length of the wind turbine and the aerodynamic response time at the rated wind speed.

6. The method according to claim 1, characterized in that, Based on the current lead time, the forward wind speed signal is delayed to generate a predicted wind speed, and a feedforward pitch command is generated based on the predicted wind speed, specifically including: The wind speed signal at the farthest point of the forward lidar is selected as the forward wind speed signal. The forward wind speed signal is delayed by the current lead time and used as the predicted value of the future wind speed on the impeller surface to obtain the predicted wind speed. Based on the predicted wind speed, current generator speed and current pitch angle, the required pitch command is calculated and superimposed on the output of the original feedback pitch controller. When the lidar data used to acquire the feature parameters fails, it automatically switches to pure feedback pitch mode. When the actual response delay exceeds a preset reasonable range, the maximum lead time is limited and a maintenance alarm is triggered.

7. The method according to claim 5, characterized in that, Based on the unit's operating status indicators after pitch control, the calculation parameters for the current lead time are corrected through feedback, specifically including: The maximum overshoot of generator speed, the time to recover to a stable speed, and the fatigue cost of pitch control after pitch adjustment are collected as the operating status indicators of the unit. When the maximum overshoot of the generator speed is greater than the first preset threshold, or the time for the speed to recover to a stable state is greater than the second preset threshold, it is determined to be an under-lead state. When the pitch control frequency is higher than the historical average and the maximum overshoot of the generator speed is less than the third preset threshold, it is determined to be an over-leading state. Based on the under-leading state or the over-leading state, the first proportional coefficient, the second proportional coefficient, the third proportional coefficient, and / or the reference lead time are adjusted in a closed loop.

8. An adaptive advance pitch control system for a wind turbine generator, characterized in that, include: The wind condition sensing module is used to acquire characteristic parameters that characterize the trend of wind speed changes in front of the wind turbine. The delay identification module is used to identify the actual response delay of the pitch actuator based on the timing difference between the pitch command issuance time and the actual pitch angle change time. The time calculation module is used to calculate the current lead time based on the characteristic parameters and the actual response delay; The instruction generation module is used to perform delay processing on the forward wind speed signal based on the current lead time to generate a predicted wind speed, and generate a feedforward pitch instruction based on the predicted wind speed. The feedback correction module is used to correct the calculation parameters of the current lead time based on the unit's operating status indicators after the pitch control operation.

9. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When a processor executes a program stored in a memory, it implements the steps of the adaptive advance pitch control method for the wind turbine generator as described in any one of claims 1-7.

10. A computer storage medium, characterized in that, The computer storage medium stores a computer program, which, when executed by a processor, implements the steps of the adaptive advance pitch control method for any of the wind turbines described in claims 1-7.

Images