Method and system for optimizing generating capacity of fan based on IPC on-load yaw strategy

Abstract

The invention relates to the technical field of wind power generation. The invention provides a method and system for optimizing the generating capacity of a fan based on an IPC on-load yaw strategy. The method comprises the following steps: collecting wind direction data in real time, and predicting a wind direction change trend in a future preset time period in combination with historical wind direction data and real-time wind direction data; according to the prediction result of the wind direction change trend and the current cabin orientation of the draught fan, the prediction deviation angle of the cabin and the wind direction is calculated, and meanwhile a dynamic yaw starting threshold value is set; the operation parameter data of the fan are monitored in real time, the yaw speed and the motor output power are dynamically adjusted according to the monitored data, the pitch angle of the wind wheel is synchronously adjusted, and the wind load in the plane of the wind wheel is balanced; and optimizing a dynamic yaw starting threshold value and a yaw adjusting parameter according to the operation data in the yaw process. The problems that due to the fact that a traditional yaw control strategy cannot track instantaneous wind direction changes in real time, wind alignment is not accurate, fatigue damage of fan components is caused, hardware abrasion is accelerated, electric quantity loss is caused, and operation and maintenance cost is increased are solved.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, and more specifically, to a method and system for optimizing wind turbine power generation based on IPC load yaw strategy. Background Technology

[0002] In wind power systems, the yaw system is a key component for achieving efficient turbine operation. Its core function is to adjust the nacelle orientation to ensure the rotor plane always faces the wind direction, maximizing wind energy capture efficiency. A traditional yaw system mainly consists of a yaw motor, yaw reduction gearbox, yaw bearing, yaw drive gearbox, and yaw brake. Its working principle is as follows: the yaw motor provides power, which drives the pinion gear through the reduction gearbox to mesh with the large gear ring of the yaw bearing fixed to the tower flange, driving the nacelle to rotate; in the non-yaw state, the yaw brake maintains nacelle stability. The yaw control system, on the other hand, collects wind direction data in real time and drives the motor to adjust the nacelle position, aiming to control the average angular deviation between the nacelle and the wind direction towards zero.

[0003] Traditional yaw control strategies rely on historical wind direction averages for decision-making, failing to track instantaneous wind direction changes in real time. This leads to significant deviations between the nacelle and the wind direction, directly causing wind load imbalance within the rotor plane. This exacerbates fatigue damage to components such as blades and hubs, significantly impacting the turbine's lifespan and operational reliability. To compensate for this misalignment, traditional strategies often adjust the nacelle direction by increasing yaw frequency. However, this introduces new problems. Frequent yaw accelerates wear and even failure of hardware components such as yaw motors, gearboxes, and friction plates. Furthermore, the turbine must be shut down and the propellers retracted during yaw; statistics show that such shutdowns result in approximately 1% of the annual electricity loss for wind farms. In addition, the impact loads generated during yaw trigger alarms in the turbine vibration detection system, further increasing downtime and maintenance costs. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for optimizing wind turbine power generation based on IPC-based load yaw strategy. This aims to solve the problem that traditional yaw control strategies cannot track instantaneous wind direction changes in real time, resulting in inaccurate wind alignment, which in turn causes fatigue damage to wind turbine components, accelerated hardware wear, power loss, and increased operation and maintenance costs.

[0005] This invention is achieved through the following technical solution: A method for optimizing wind turbine power generation based on IPC-based load yaw strategy includes the following steps: Real-time wind direction data is collected, and combined with historical and real-time wind direction data, the trend of wind direction changes within a preset time period is predicted. Based on the predicted wind direction change trend and the current nacelle orientation of the wind turbine, the predicted deviation angle between the nacelle and the wind direction is calculated, and the dynamic yaw start threshold is set under different wind direction change trends and wind speed conditions. When the predicted deviation angle is greater than the corresponding dynamic yaw start threshold, the load yaw command is triggered. Under the continuous power generation operation of the wind turbine, the yaw motor is controlled to drive the yaw reduction gearbox, yaw drive gearbox and yaw bearing to work together to drive the nacelle to rotate in the wind direction. During the yaw process with load, the operating parameters of the wind turbine are monitored in real time, and the yaw speed and motor output power are dynamically adjusted according to the monitoring data. The rotor pitch angle is also adjusted synchronously to balance the wind load in the rotor plane. When the deviation angle between the nacelle and the wind direction is less than the set yaw end threshold, the yaw motor stops working, the yaw braking device is activated and the nacelle position is fixed, and the yaw adjustment is completed. After yaw is completed, the dynamic yaw initiation threshold and yaw adjustment parameters are optimized based on the operational data during the yaw process, providing an optimization strategy for subsequent yaw control.

[0006] Optionally, the specific process of predicting the wind direction change trend within a preset time period by collecting real-time wind direction data, combining historical wind direction data, and current real-time wind direction data is as follows: Wind direction data is collected in real time by a wind direction sensor installed on the top of the cabin, and wind direction time series data within a preset historical period is retrieved simultaneously from the SCADA system. The collected real-time wind direction data and historical data are preprocessed to obtain a preprocessed dataset; Based on the preprocessed dataset, a hybrid prediction model integrating time series prediction model and machine learning algorithm is constructed. The input is the current wind speed, turbulence intensity and historical wind direction change pattern, and the output is the wind direction change curve within a preset time period in the future. Based on the wind direction change curve, the wind direction change rate and directional stability index are calculated, the confidence interval of the prediction results are dynamically corrected, and a wind direction change trend prediction result with time series characteristics is generated.

[0007] Optionally, the specific process of constructing a hybrid prediction model that integrates a time series prediction model and a machine learning algorithm based on a preprocessed dataset is as follows: Preprocessed wind direction time series data Input the ARIMA model and construct the basic wind direction prediction equation as shown in equation (1):

[0008] in, express The order difference operator is used to stationary time series data. Indicates time The actual wind direction angle; representing a constant term; This represents the coefficient of the autoregressive term, reflecting the historical wind direction angle. Impact on the current value; Indicates the order of autoregression; This represents the moving average coefficient, reflecting historical noise. Impact on the current value; Indicates the order of the moving average; time White noise error; output future Baseline wind direction forecast for the period ; Constructing feature vectors ;in, Indicates real-time wind speed; Indicates turbulence intensity; Representing history Time-series of wind direction changes; Indicates that the ARIMA model is in moment to moment The predicted wind direction; The feature vectors are input into the gradient boosting decision tree model to learn the time series model residuals. , Indicates that the ARIMA model is in moment to moment The wind direction forecast is used to generate a corrected forecast, as shown in equation (2) below:

[0009] in, This represents the wind direction prediction value after correction by the gradient boosting decision tree; This represents the prediction function of a gradient boosting decision tree. Indicates that the ARIMA model is in moment to moment The predicted wind direction; Based on the current wind speed range and turbulence intensity level, the fusion weighting factor is calculated as shown in equation (3) below:

[0010] in, The fusion weighting factor represents the fusion weighting of machine learning prediction results; This represents the weight adjustment coefficient, used to control the sensitivity of the weight to changes in turbulence intensity; Indicates the turbulence intensity threshold; The final mixed prediction output is generated as shown in equation (4) below:

[0011] in, This represents the output of the final hybrid model. Forecast wind direction at any time; This represents the predicted value corrected by the gradient boosting decision tree.

[0012] Optionally, the specific process of calculating the wind direction change rate and directional stability index based on the wind direction change curve, dynamically correcting the confidence interval of the prediction result, and generating a wind direction change trend prediction result with time-series characteristics is as follows: The wind direction change curve output by the hybrid prediction model is subjected to time series differentiation processing to calculate the instantaneous wind direction change rate at each sampling point within a preset future time period. ; Based on the wind direction change curve, the standard deviation of the wind direction angle is calculated within a preset sliding time window. and standard deviation As an indicator of directional stability, it reflects the magnitude of wind direction fluctuations; Combined with wind direction change rate and directional stability index Adjust the confidence interval width of the predicted value according to the following rules: when or At that time, expand the confidence interval to ; when and At that time, narrow the confidence interval to ; in, and This is a preset turbulence threshold; and is the interval width coefficient, and ; The corrected confidence interval, wind direction change rate, and directional stability index are bound to the original wind direction prediction value to generate a structured time-series feature data body containing timestamps, predicted wind direction, confidence interval boundaries, change rate, and stability index. This data serves as the risk change trend prediction result and is output to the yaw control module.

[0013] Optionally, the specific process of monitoring the wind turbine's operating parameters in real time during the loaded yaw process, dynamically adjusting the yaw speed and motor output power based on the monitoring data, and simultaneously adjusting the rotor pitch angle to balance the wind load in the rotor plane is as follows: The SCADA system collects the wind turbine's operating parameters in real time, including wind speed, generator power, nacelle vibration acceleration, blade root bending moment, and yaw bearing stress distribution data. Based on the predicted deviation angle and real-time wind speed, the basic yaw speed is calculated according to a preset nonlinear function, as shown in equation (5) below:

[0014] in, Indicates the target's yaw speed; Indicates the angle of prediction deviation; Indicates real-time wind speed; Indicates the speed reference coefficient; Indicates the angle sensitivity factor; Indicates the wind speed attenuation index; The yaw speed is corrected in real time based on the difference in blade root bending moment, when the bending moment difference between any two blades is... At that time, a deceleration compensation amount is generated. ;in, Indicates the threshold for bending moment difference; and This represents the bending moment between any two blades; Indicates the maximum bending moment gradient; Indicates the compensation gain coefficient; Based on the base yaw speed and the deceleration compensation, the final executed yaw speed is output as shown in equation (6):

[0015] in, Indicates the final yaw speed; and These represent the system's preset minimum and maximum safety limits, respectively. The independent pitch control strategy is started synchronously, and the wind turbine planar wind load imbalance is calculated as shown in the following formula (7):

[0016] in, Indicates the standard deviation of the leaf root bending moment; , and These correspond to the bending moments at the roots of the three blades of the wind turbine. This represents the average bending moment of the three blades; Indicates the prevention of zero constant; The independent pitch angle compensation for each blade is generated as shown in equation (8):

[0017] in, This indicates the amount of independent pitch angle compensation for the blades; Indicates the pitch gain coefficient; Indicates the blade azimuth angle; indicates the nacelle yaw angle offset. Based on cabin vibration acceleration feedback, the motor output power is adjusted in a closed loop when the vibration amplitude... At that time, the motor power is dynamically limited according to the following formula (8):

[0018] in, This indicates the actual yaw motor power being used; Indicates the rated drive power; Indicates the vibration suppression coefficient; This indicates the amplitude of the cabin vibration acceleration as monitored in real time. Indicates the cabin vibration safety threshold; Continuously monitor the rate of change of wind load imbalance during the yaw process, when At that time, it is determined that the wind load has returned to balance; among them, This is the convergence threshold.

[0019] Optionally, after the yaw is completed, the specific process of optimizing the dynamic yaw initiation threshold and yaw adjustment parameters based on the operational data during this yaw process is as follows: The operational data during this yaw process were normalized to construct a yaw operation feature dataset. Based on the yaw operation feature dataset, a clustering analysis algorithm is used to classify yaw conditions under different wind direction change trends and different wind speeds. The mean deviation between the predicted deviation angle and the actual yaw angle under each condition is calculated. When the mean deviation exceeds the preset dynamic threshold adjustment sensitivity, the dynamic yaw start threshold is optimized according to the following rules: if the mean deviation is positive, the dynamic yaw start threshold under the corresponding condition is reduced proportionally; if the mean deviation is negative, the dynamic yaw start threshold under the corresponding condition is increased proportionally. A genetic algorithm is used to globally optimize the yaw adjustment parameters. The optimization objectives are to minimize the wind farm's power loss and the wear of the yaw components during the yaw process. The optimized yaw adjustment parameter combination is generated by combining the corresponding parameters in the operation data. The optimized dynamic yaw initiation threshold and yaw adjustment parameter combination are fed back into the hybrid prediction model that integrates the time series prediction model and the machine learning algorithm to update the model parameters and provide a more accurate strategic basis for subsequent wind direction change trend prediction and yaw control.

[0020] Optionally, the specific process of using a genetic algorithm to globally optimize the yaw adjustment parameters is as follows: The variables related to yaw adjustment parameters in the yaw operation feature dataset are encoded to form an initial population of individuals, with each individual corresponding to a set of yaw adjustment parameter combinations; A fitness function was constructed with the optimization objectives of minimizing wind farm power loss and minimizing wear of yaw components during this yaw process. The fitness values ​​of each individual in the initial population were calculated by combining parameters such as power generation, yaw motor operating time, and gearbox wear from the operating data. Based on fitness values, individuals with high fitness are selected from the initial population through selection operations. Crossover operations are then used to recombine the genes of the selected individuals to generate new individuals. Mutation operations are then performed on the new individuals to introduce genetic diversity. Repeatedly perform selection, crossover, and mutation operations to generate the next generation population, calculate the fitness value of each individual in the next generation population, and determine whether the preset termination conditions are met. The termination conditions include reaching the maximum number of iterations and the population fitness value converging to a preset precision. When the termination condition is met, the yaw adjustment parameter combination corresponding to the individual with the highest fitness value is used as the optimized yaw adjustment parameter combination.

[0021] Based on the same inventive concept, this invention also provides a method for optimizing wind turbine power generation based on IPC load-bearing yaw strategy, comprising: The data acquisition module is used to collect wind direction data in real time through a wind direction sensor installed on the top of the nacelle, and retrieve wind direction time series data within a preset historical period from the SCADA system, while also collecting wind turbine operating parameters; The wind direction change trend prediction module is used to preprocess the collected real-time wind direction data and historical data. Based on the preprocessed dataset, it constructs a hybrid prediction model that integrates time series prediction model and machine learning algorithm. The input is the current wind speed, turbulence intensity and historical wind direction change pattern. The output is the wind direction change curve within the future preset time period. The wind direction change rate and direction stability index are calculated. The confidence interval of the prediction results is dynamically corrected to generate wind direction change trend prediction results with time series characteristics. The yaw decision module is used to calculate the predicted deviation angle between the nacelle and the wind direction based on the predicted wind direction change trend and the current nacelle orientation of the wind turbine. At the same time, it sets the dynamic yaw start threshold under different wind direction change trends and wind speed conditions. When the predicted deviation angle is greater than the corresponding dynamic yaw start threshold, the load yaw command is triggered. The load-bearing yaw execution control module is used to control the yaw motor to drive the yaw reduction gearbox, yaw drive gearbox and yaw bearing to work together in the continuous power generation state of the wind turbine after receiving the load-bearing yaw command. This drives the nacelle to rotate in the wind direction. During the rotation, the module monitors the wind turbine's operating parameters in real time, dynamically adjusts the yaw speed and motor output power based on the monitoring data, and synchronously adjusts the wind turbine pitch angle to balance the wind load in the wind turbine plane. The yaw termination module is used to control the yaw motor to stop working, activate the yaw braking device and fix the nacelle position when the deviation angle between the nacelle and the wind direction is less than the set yaw termination threshold, thus completing the yaw adjustment. The parameter optimization module is used to normalize the operational data during this yaw process, construct a yaw operation feature dataset, optimize the dynamic yaw initiation threshold using a clustering analysis algorithm based on this dataset, and perform global optimization of the yaw adjustment parameters using a genetic algorithm. The optimized dynamic yaw initiation threshold and yaw adjustment parameter combination are then fed back into the hybrid prediction model to update the model parameters.

[0022] Based on the same inventive concept, the present invention also provides an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the above-described method for optimizing wind turbine power generation based on IPC load yaw strategy.

[0023] Based on the same inventive concept, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the above-described method for optimizing wind turbine power generation based on IPC load yaw strategy.

[0024] The technical solution of the present invention has at least the following advantages and beneficial effects: By collecting wind direction data in real time and combining it with historical data to predict future wind direction trends, compared with the traditional yaw control strategy that makes decisions based on the average historical wind direction, it can more accurately predict wind direction and calculate the nacelle's deviation angle from the wind direction prediction more accurately. This allows the yaw system to adjust the nacelle's direction in a timely and accurate manner, reducing the wind load imbalance problem in the wind turbine plane caused by inaccurate wind alignment, maximizing wind energy capture efficiency, and effectively avoiding power generation loss caused by untimely wind direction tracking, thereby significantly improving the power generation efficiency of wind turbines.

[0025] By setting dynamic yaw initiation thresholds under different wind direction and speed conditions, the yaw command is triggered only when the predicted deviation angle is greater than the corresponding threshold. This avoids the frequent yaw caused by inaccurate wind alignment in traditional strategies. Reducing the frequency of yaw can effectively reduce the wear rate of hardware components such as yaw motors, gearboxes, and friction plates, extend their service life, reduce the probability of hardware failure, and reduce equipment replacement and maintenance costs.

[0026] This technology enables yaw under load while the wind turbine is continuously generating electricity, changing the traditional mode where the wind turbine needs to be stopped and the propeller retracted during yaw. This avoids the annual power loss of the wind farm caused by traditional shutdown yaw and further increases the annual power generation of the wind farm.

[0027] During yaw under load, the wind turbine's operating parameters are monitored in real time, and the yaw speed and motor output power are dynamically adjusted based on the monitoring data. At the same time, the rotor pitch angle is adjusted synchronously to balance the wind load in the rotor plane. This can effectively reduce the impact load generated during yaw, avoid triggering the wind turbine vibration detection system alarm, reduce the additional downtime caused by alarms, ensure the stability of wind turbine operation, and reduce operation and maintenance costs.

[0028] After yaw is completed, the dynamic yaw initiation threshold and yaw adjustment parameters are optimized based on the operational data during the yaw process. By continuously accumulating and analyzing operational data, the yaw control strategy can be continuously optimized according to actual operating conditions, thereby achieving more efficient and precise yaw control under different environmental conditions, and further improving the overall performance and operational reliability of the wind turbine. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating the method for optimizing wind turbine power generation based on IPC load yaw strategy according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the system for optimizing wind turbine power generation based on IPC load yaw strategy according to an embodiment of the present invention. Detailed Implementation

[0030] The following is a detailed description of the embodiments, in conjunction with the accompanying drawings.

[0031] Reference Figure 1 A method for optimizing wind turbine power generation based on IPC-based load yaw strategy includes the following steps: Step 1: Collect wind direction data in real time, and combine historical wind direction data with real-time wind direction data to predict the wind direction change trend within a preset time period.

[0032] In some embodiments, the specific process of collecting wind direction data in real time, combining historical wind direction data and current real-time wind direction data, and predicting the wind direction change trend within a preset time period is as follows: Wind direction data is collected in real time by a wind direction sensor installed on the top of the cabin, and wind direction time series data within a preset historical period is retrieved simultaneously from the SCADA system. The collected real-time wind direction data and historical data are preprocessed to obtain a preprocessed dataset; Based on the preprocessed dataset, a hybrid prediction model integrating time series prediction model and machine learning algorithm is constructed. The input is the current wind speed, turbulence intensity and historical wind direction change pattern, and the output is the wind direction change curve within a preset time period in the future. Based on the wind direction change curve, the wind direction change rate and directional stability index are calculated, the confidence interval of the prediction results are dynamically corrected, and a wind direction change trend prediction result with time series characteristics is generated.

[0033] In some embodiments, the specific process of constructing a hybrid prediction model that integrates a time series prediction model and a machine learning algorithm based on a preprocessed dataset is as follows: Preprocessed wind direction time series data Input the ARIMA model and construct the basic wind direction prediction equation as shown in equation (1):

[0034] in, express The order difference operator is used to stationary time series data. Indicates time The actual wind direction angle; representing a constant term; This represents the coefficient of the autoregressive term, reflecting the historical wind direction angle. Impact on the current value; Indicates the order of autoregression; This represents the moving average coefficient, reflecting historical noise. Impact on the current value; Indicates the order of the moving average; time White noise error; output future Baseline wind direction forecast for the period ; Constructing feature vectors ;in, Indicates real-time wind speed; Indicates turbulence intensity; Representing history Time-series of wind direction changes; Indicates that the ARIMA model is in moment to moment The predicted wind direction; The feature vectors are input into the gradient boosting decision tree model to learn the time series model residuals. , Indicates that the ARIMA model is in moment to moment The wind direction forecast is used to generate a corrected forecast, as shown in equation (2) below:

[0035] in, This represents the wind direction prediction value after correction by the gradient boosting decision tree; This represents the prediction function of a gradient boosting decision tree. Indicates that the ARIMA model is in moment to moment The predicted wind direction; Based on the current wind speed range and turbulence intensity level, the fusion weighting factor is calculated as shown in equation (3) below:

[0036] in, The fusion weighting factor represents the fusion weighting of machine learning prediction results; This represents the weight adjustment coefficient, used to control the sensitivity of the weight to changes in turbulence intensity; Indicates the turbulence intensity threshold; The final mixed prediction output is generated as shown in equation (4) below:

[0037] in, This represents the output of the final hybrid model. Forecast wind direction at any time; This represents the predicted value corrected by the gradient boosting decision tree.

[0038] In some embodiments, the specific process of calculating the wind direction change rate and directional stability index based on the wind direction change curve, dynamically correcting the confidence interval of the prediction results, and generating a wind direction change trend prediction result with time-series characteristics is as follows: The wind direction change curve output by the hybrid prediction model is subjected to time series differentiation processing to calculate the instantaneous wind direction change rate at each sampling point within a preset future time period. ; Based on the wind direction change curve, the standard deviation of the wind direction angle is calculated within a preset sliding time window. and standard deviation As an indicator of directional stability, it reflects the magnitude of wind direction fluctuations; Combined with wind direction change rate and directional stability index Adjust the confidence interval width of the predicted value according to the following rules: when or At that time, expand the confidence interval to ; when and At that time, narrow the confidence interval to ; in, and This is a preset turbulence threshold; and is the interval width coefficient, and ; The corrected confidence interval, wind direction change rate, and directional stability index are bound to the original wind direction prediction value to generate a structured time-series feature data body containing timestamps, predicted wind direction, confidence interval boundaries, change rate, and stability index. This data serves as the risk change trend prediction result and is output to the yaw control module.

[0039] Step 2: Based on the predicted wind direction trend and the current nacelle orientation of the wind turbine, calculate the predicted deviation angle between the nacelle and the wind direction, and set the dynamic yaw start threshold under different wind direction trends and wind speed conditions.

[0040] In some embodiments, the current orientation angle of the cabin is collected in real time by an azimuth sensor (such as an encoder or GPS positioning device) installed at the bottom of the cabin. This angle is taken as 0° north and rotated clockwise to the centerline of the cabin. From the structured time-series feature data volume output in step one, the predicted wind direction value for a preset time period in the future is extracted. This data includes a timestamp, the mean of the predicted wind direction, and the confidence interval boundary.

[0041] Step 3: When the predicted deviation angle is greater than the corresponding dynamic yaw start threshold, the load yaw command is triggered. Under the continuous power generation operation of the wind turbine, the yaw motor is controlled to drive the yaw reduction gearbox, yaw drive gearbox and yaw bearing to work together to drive the nacelle to rotate in the wind direction. Step 4: During the yaw process under load, monitor the operating parameters of the wind turbine in real time, dynamically adjust the yaw speed and motor output power based on the monitoring data, and simultaneously adjust the rotor pitch angle to balance the wind load in the rotor plane.

[0042] In some embodiments, the specific process of monitoring the wind turbine's operating parameters in real time during yaw under load, dynamically adjusting the yaw speed and motor output power based on the monitoring data, and simultaneously adjusting the rotor pitch angle to balance the wind load in the rotor plane is as follows: The SCADA system collects the wind turbine's operating parameters in real time, including wind speed, generator power, nacelle vibration acceleration, blade root bending moment, and yaw bearing stress distribution data. Based on the predicted deviation angle and real-time wind speed, the basic yaw speed is calculated according to a preset nonlinear function, as shown in equation (5) below:

[0043] in, Indicates the target's yaw speed; Indicates the angle of prediction deviation; Indicates real-time wind speed; Indicates the speed reference coefficient; Indicates the angle sensitivity factor; Indicates the wind speed attenuation index; The yaw speed is corrected in real time based on the difference in blade root bending moment, when the bending moment difference between any two blades is... At that time, a deceleration compensation amount is generated. ;in, Indicates the threshold for bending moment difference; and This represents the bending moment between any two blades; Indicates the maximum bending moment gradient; Indicates the compensation gain coefficient; Based on the base yaw speed and the deceleration compensation, the final executed yaw speed is output as shown in equation (6):

[0044] in, Indicates the final yaw speed; and These represent the system's preset minimum and maximum safety limits, respectively. The independent pitch control strategy is started synchronously, and the wind turbine planar wind load imbalance is calculated as shown in the following formula (7):

[0045] in, Indicates the standard deviation of the leaf root bending moment; , and These correspond to the bending moments at the roots of the three blades of the wind turbine. This represents the average bending moment of the three blades; Indicates the prevention of zero constant; The independent pitch angle compensation for each blade is generated as shown in equation (8):

[0046] in, This indicates the amount of independent pitch angle compensation for the blades; Indicates the pitch gain coefficient; Indicates the blade azimuth angle; indicates the nacelle yaw angle offset. Based on cabin vibration acceleration feedback, the motor output power is adjusted in a closed loop when the vibration amplitude... At that time, the motor power is dynamically limited according to the following formula (8):

[0047] in, This indicates the actual yaw motor power being used; Indicates the rated drive power; Indicates the vibration suppression coefficient; This indicates the amplitude of the cabin vibration acceleration as monitored in real time. Indicates the cabin vibration safety threshold; Continuously monitor the rate of change of wind load imbalance during the yaw process, when At that time, it is determined that the wind load has returned to balance; among them, This is the convergence threshold.

[0048] Step 5: When the deviation angle between the nacelle and the wind direction is less than the set yaw end threshold, the yaw motor stops working, the yaw braking device is activated and the nacelle position is fixed, and the yaw adjustment is completed. Step 6: After yaw is completed, optimize the dynamic yaw initiation threshold and yaw adjustment parameters based on the operational data during this yaw process, providing an optimization strategy for subsequent yaw control.

[0049] In some embodiments, after the yaw is completed, the specific process of optimizing the dynamic yaw initiation threshold and yaw adjustment parameters based on the operational data during this yaw process is as follows: The operational data during this yaw process were normalized to construct a yaw operation feature dataset. Based on the yaw operation feature dataset, a clustering analysis algorithm is used to classify yaw conditions under different wind direction change trends and different wind speeds. The mean deviation between the predicted deviation angle and the actual yaw angle under each condition is calculated. When the mean deviation exceeds the preset dynamic threshold adjustment sensitivity, the dynamic yaw start threshold is optimized according to the following rules: if the mean deviation is positive, the dynamic yaw start threshold under the corresponding condition is reduced proportionally; if the mean deviation is negative, the dynamic yaw start threshold under the corresponding condition is increased proportionally. A genetic algorithm is used to globally optimize the yaw adjustment parameters. The optimization objectives are to minimize the wind farm's power loss and the wear of the yaw components during the yaw process. The optimized yaw adjustment parameter combination is generated by combining the corresponding parameters in the operation data. The optimized dynamic yaw initiation threshold and yaw adjustment parameter combination are fed back into the hybrid prediction model that integrates the time series prediction model and the machine learning algorithm to update the model parameters and provide a more accurate strategic basis for subsequent wind direction change trend prediction and yaw control.

[0050] In some embodiments, the specific process of using a genetic algorithm to globally optimize the yaw adjustment parameters is as follows: The variables related to yaw adjustment parameters in the yaw operation feature dataset are encoded to form an initial population of individuals, with each individual corresponding to a set of yaw adjustment parameter combinations; A fitness function was constructed with the optimization objectives of minimizing wind farm power loss and minimizing wear of yaw components during this yaw process. The fitness values ​​of each individual in the initial population were calculated by combining parameters such as power generation, yaw motor operating time, and gearbox wear from the operating data. Based on fitness values, individuals with high fitness are selected from the initial population through selection operations. Crossover operations are then used to recombine the genes of the selected individuals to generate new individuals. Mutation operations are then performed on the new individuals to introduce genetic diversity. Repeatedly perform selection, crossover, and mutation operations to generate the next generation population, calculate the fitness value of each individual in the next generation population, and determine whether the preset termination conditions are met. The termination conditions include reaching the maximum number of iterations and the population fitness value converging to a preset precision. When the termination condition is met, the yaw adjustment parameter combination corresponding to the individual with the highest fitness value is used as the optimized yaw adjustment parameter combination.

[0051] Based on the same inventive concept, and corresponding to any of the above embodiments, refer to... Figure 2 This invention provides a system for optimizing wind turbine power generation based on IPC-based load yaw strategy, used to implement the aforementioned method for optimizing wind turbine power generation based on IPC-based load yaw strategy, comprising: The data acquisition module is used to collect wind direction data in real time through a wind direction sensor installed on the top of the nacelle, and retrieve wind direction time series data within a preset historical period from the SCADA system, while also collecting wind turbine operating parameters; The wind direction change trend prediction module is used to preprocess the collected real-time wind direction data and historical data. Based on the preprocessed dataset, it constructs a hybrid prediction model that integrates time series prediction model and machine learning algorithm. The input is the current wind speed, turbulence intensity and historical wind direction change pattern. The output is the wind direction change curve within the future preset time period. The wind direction change rate and direction stability index are calculated. The confidence interval of the prediction results is dynamically corrected to generate wind direction change trend prediction results with time series characteristics. The yaw decision module is used to calculate the predicted deviation angle between the nacelle and the wind direction based on the predicted wind direction change trend and the current nacelle orientation of the wind turbine. At the same time, it sets the dynamic yaw start threshold under different wind direction change trends and wind speed conditions. When the predicted deviation angle is greater than the corresponding dynamic yaw start threshold, the load yaw command is triggered. The load-bearing yaw execution control module is used to control the yaw motor to drive the yaw reduction gearbox, yaw drive gearbox and yaw bearing to work together in the continuous power generation state of the wind turbine after receiving the load-bearing yaw command. This drives the nacelle to rotate in the wind direction. During the rotation, the module monitors the wind turbine's operating parameters in real time, dynamically adjusts the yaw speed and motor output power based on the monitoring data, and synchronously adjusts the wind turbine pitch angle to balance the wind load in the wind turbine plane. The yaw termination module is used to control the yaw motor to stop working, activate the yaw braking device and fix the nacelle position when the deviation angle between the nacelle and the wind direction is less than the set yaw termination threshold, thus completing the yaw adjustment. The parameter optimization module is used to normalize the operational data during this yaw process, construct a yaw operation feature dataset, optimize the dynamic yaw initiation threshold using a clustering analysis algorithm based on this dataset, and perform global optimization of the yaw adjustment parameters using a genetic algorithm. The optimized dynamic yaw initiation threshold and yaw adjustment parameter combination are then fed back into the hybrid prediction model to update the model parameters.

[0052] Based on the same inventive concept, corresponding to any of the above embodiments, the present invention provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the method of optimizing wind turbine power generation based on IPC load yaw strategy of the embodiment.

[0053] Alternatively, the aforementioned electronic device may be a server.

[0054] In addition, this embodiment also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method of optimizing wind turbine power generation based on IPC load yaw strategy of the embodiment.

[0055] It is understood that the processor in the embodiments of the present invention can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor can be a microprocessor or any conventional processor.

[0056] The method steps in the embodiments of the present invention can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0057] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a storage medium or transmitted through a storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

Claims

1. A method for optimizing wind turbine power generation based on IPC-based load yaw strategy, characterized in that, Includes the following steps: Real-time wind direction data is collected, and combined with historical and real-time wind direction data, the trend of wind direction changes within a preset time period is predicted. Based on the predicted wind direction change trend and the current nacelle orientation of the wind turbine, the predicted deviation angle between the nacelle and the wind direction is calculated, and the dynamic yaw start threshold is set under different wind direction change trends and wind speed conditions. When the predicted deviation angle is greater than the corresponding dynamic yaw start threshold, the load yaw command is triggered. Under the continuous power generation operation of the wind turbine, the yaw motor is controlled to drive the yaw reduction gearbox, yaw drive gearbox and yaw bearing to work together to drive the nacelle to rotate in the wind direction. During the yaw process with load, the operating parameters of the wind turbine are monitored in real time, and the yaw speed and motor output power are dynamically adjusted according to the monitoring data. The rotor pitch angle is also adjusted synchronously to balance the wind load in the rotor plane. When the deviation angle between the nacelle and the wind direction is less than the set yaw end threshold, the yaw motor stops working, the yaw braking device is activated and the nacelle position is fixed, and the yaw adjustment is completed. After yaw is completed, the dynamic yaw initiation threshold and yaw adjustment parameters are optimized based on the operational data during the yaw process, providing an optimization strategy for subsequent yaw control.

2. The method for optimizing wind turbine power generation based on IPC-based load yaw strategy as described in claim 1, characterized in that, The specific process of collecting real-time wind direction data, combining historical wind direction data and current real-time wind direction data, to predict the wind direction change trend within a preset time period is as follows: Wind direction data is collected in real time by a wind direction sensor installed on the top of the cabin, and wind direction time series data within a preset historical period is retrieved simultaneously from the SCADA system. The collected real-time wind direction data and historical data are preprocessed to obtain a preprocessed dataset; Based on the preprocessed dataset, a hybrid prediction model integrating time series prediction model and machine learning algorithm is constructed. The input is the current wind speed, turbulence intensity and historical wind direction change pattern, and the output is the wind direction change curve within a preset time period in the future. Based on the wind direction change curve, the wind direction change rate and directional stability index are calculated, the confidence interval of the prediction results are dynamically corrected, and a wind direction change trend prediction result with time series characteristics is generated.

3. The method for optimizing wind turbine power generation based on IPC-based load yaw strategy as described in claim 2, characterized in that, The specific process of constructing a hybrid prediction model that integrates a time series prediction model and a machine learning algorithm based on a preprocessed dataset is as follows: Preprocessed wind direction time series data Input the ARIMA model and construct the basic wind direction prediction equation as shown in equation (1): in, express The order difference operator is used to stationary time series data. Indicates time The actual wind direction angle; representing a constant term; This represents the coefficient of the autoregressive term, reflecting the historical wind direction angle. Impact on the current value; Indicates the order of autoregression; This represents the moving average coefficient, reflecting historical noise. Impact on the current value; Indicates the order of the moving average; time White noise error; output future Baseline wind direction forecast for the period ; Constructing feature vectors ;in, Indicates real-time wind speed; Indicates turbulence intensity; Representing history Time-series of wind direction changes; Indicates that the ARIMA model is in moment to moment The predicted wind direction; The feature vectors are input into the gradient boosting decision tree model to learn the time series model residuals. , Indicates that the ARIMA model is in moment to moment The wind direction forecast is used to generate a corrected forecast, as shown in equation (2) below: in, This represents the wind direction prediction value after correction by the gradient boosting decision tree; This represents the prediction function of a gradient boosting decision tree. Indicates that the ARIMA model is in moment to moment The predicted wind direction; Based on the current wind speed range and turbulence intensity level, the fusion weighting factor is calculated as shown in equation (3) below: in, The fusion weighting factor represents the fusion weighting of machine learning prediction results; This represents the weight adjustment coefficient, used to control the sensitivity of the weight to changes in turbulence intensity; Indicates the turbulence intensity threshold; The final mixed prediction output is generated as shown in equation (4) below: in, This represents the output of the final hybrid model. Forecast wind direction at any time; This represents the predicted value corrected by the gradient boosting decision tree.

4. The method for optimizing wind turbine power generation based on IPC load yaw strategy as described in claim 2, characterized in that, The specific process of calculating the wind direction change rate and directional stability index based on the wind direction change curve, dynamically correcting the confidence interval of the prediction results, and generating a wind direction change trend prediction result with time-series characteristics is as follows: The wind direction change curve output by the hybrid prediction model is subjected to time series differentiation processing to calculate the instantaneous wind direction change rate at each sampling point within a preset future time period. ; Based on the wind direction change curve, the standard deviation of the wind direction angle is calculated within a preset sliding time window. and standard deviation As an indicator of directional stability, it reflects the magnitude of wind direction fluctuations; Combined with wind direction change rate and directional stability index Adjust the confidence interval width of the predicted value according to the following rules: when or At that time, expand the confidence interval to ; when and When, narrow the confidence interval to ; in, and This is a preset turbulence threshold; and is the interval width coefficient, and ; The corrected confidence interval, wind direction change rate, and directional stability index are bound to the original wind direction prediction value to generate a structured time-series feature data body containing timestamps, predicted wind direction, confidence interval boundaries, change rate, and stability index. This data body serves as the prediction result of the risk change trend and is output to the yaw control module.

5. The method for optimizing wind turbine power generation based on IPC load yaw strategy as described in claim 1, characterized in that, The specific process of monitoring the wind turbine's operating parameters in real time during the loaded yaw process, dynamically adjusting the yaw speed and motor output power based on the monitoring data, and simultaneously adjusting the turbine pitch angle to balance the wind load in the turbine plane is as follows: The SCADA system collects the wind turbine's operating parameters in real time, including wind speed, generator power, nacelle vibration acceleration, blade root bending moment, and yaw bearing stress distribution data. Based on the predicted deviation angle and real-time wind speed, the basic yaw speed is calculated according to a preset nonlinear function, as shown in equation (5) below: in, Indicates the target's yaw speed; Indicates the prediction deviation angle; Indicates real-time wind speed; Indicates the speed reference coefficient; Indicates the angle sensitivity factor; Indicates the wind speed attenuation index; The yaw speed is corrected in real time based on the difference in blade root bending moment, when the bending moment difference between any two blades is... At that time, a deceleration compensation amount is generated. ;in, Indicates the threshold for bending moment difference; and This represents the bending moment between any two blades; Indicates the maximum bending moment gradient; Indicates the compensation gain coefficient; Based on the base yaw speed and the deceleration compensation, the final executed yaw speed is output as shown in equation (6): in, Indicates the final yaw speed; and These represent the system's preset minimum and maximum safety limits, respectively. The independent pitch control strategy is started synchronously, and the wind turbine planar wind load imbalance is calculated as shown in the following formula (7): in, Indicates the standard deviation of the leaf root bending moment; , and These correspond to the bending moments at the roots of the three blades of the wind turbine. This represents the average bending moment of the three blades; Indicates the prevention of zero constant; The independent pitch angle compensation for each blade is generated as shown in equation (8): in, This indicates the amount of independent pitch angle compensation for the blades; Indicates the pitch gain coefficient; Indicates the blade azimuth angle; indicates the nacelle yaw angle offset. Based on cabin vibration acceleration feedback, the motor output power is adjusted in a closed loop when the vibration amplitude... At that time, the motor power is dynamically limited according to the following formula (8): in, This indicates the actual yaw motor power being used; Indicates the rated drive power; Indicates the vibration suppression coefficient; This represents the real-time monitored amplitude of cabin vibration acceleration. Indicates the cabin vibration safety threshold; Continuously monitor the rate of change of wind load imbalance during the yaw process, when At that time, it is determined that the wind load has returned to balance; among them, This is the convergence threshold.

6. The method for optimizing wind turbine power generation based on IPC load yaw strategy as described in claim 1, characterized in that, After the yaw is completed, the specific process of optimizing the dynamic yaw initiation threshold and yaw adjustment parameters based on the operational data during this yaw process is as follows: The operational data during this yaw process were normalized to construct a yaw operation feature dataset. Based on the yaw operation feature dataset, a clustering analysis algorithm is used to classify yaw conditions under different wind direction change trends and different wind speeds. The mean deviation between the predicted deviation angle and the actual yaw angle under each condition is calculated. When the mean deviation exceeds the preset dynamic threshold adjustment sensitivity, the dynamic yaw start threshold is optimized according to the following rules: if the mean deviation is positive, the dynamic yaw start threshold under the corresponding condition is reduced proportionally; if the mean deviation is negative, the dynamic yaw start threshold under the corresponding condition is increased proportionally. A genetic algorithm is used to globally optimize the yaw adjustment parameters. The optimization objectives are to minimize the wind farm's power loss and the wear of the yaw components during the yaw process. The optimized combination of yaw adjustment parameters is generated by combining the corresponding parameters in the operation data. The optimized dynamic yaw initiation threshold and yaw adjustment parameter combination are fed back into the hybrid prediction model that integrates the time series prediction model and the machine learning algorithm to update the model parameters and provide a more accurate strategic basis for subsequent wind direction change trend prediction and yaw control.

7. The method for optimizing wind turbine power generation based on IPC load yaw strategy as described in claim 6, characterized in that, The specific process of using a genetic algorithm to globally optimize the yaw adjustment parameters is as follows: The variables related to yaw adjustment parameters in the yaw operation feature dataset are encoded to form an initial population of individuals, with each individual corresponding to a set of yaw adjustment parameter combinations; A fitness function was constructed with the optimization objectives of minimizing wind farm power loss and minimizing wear of yaw components during this yaw process. The fitness values ​​of each individual in the initial population were calculated by combining parameters such as power generation, yaw motor operating time, and gearbox wear from the operating data. Based on fitness values, individuals with high fitness are selected from the initial population through selection operations. Crossover operations are then used to recombine the genes of the selected individuals to generate new individuals. Mutation operations are then performed on the new individuals to introduce genetic diversity. Repeatedly perform selection, crossover, and mutation operations to generate the next generation population, calculate the fitness value of each individual in the next generation population, and determine whether the preset termination conditions are met. The termination conditions include reaching the maximum number of iterations and the population fitness value converging to a preset precision. When the termination condition is met, the yaw adjustment parameter combination corresponding to the individual with the highest fitness value is used as the optimized yaw adjustment parameter combination.

8. A system for optimizing wind turbine power generation based on IPC-based load-carrying yaw strategy, used to implement the method for optimizing wind turbine power generation based on IPC-based load-carrying yaw strategy as described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to collect wind direction data in real time through a wind direction sensor installed on the top of the nacelle, and to retrieve wind direction time series data within a preset historical period from the SCADA system, while also collecting wind turbine operating parameters; The wind direction change trend prediction module is used to preprocess the collected real-time wind direction data and historical data. Based on the preprocessed dataset, it constructs a hybrid prediction model that integrates time series prediction model and machine learning algorithm. The input is the current wind speed, turbulence intensity and historical wind direction change pattern. The output is the wind direction change curve within the future preset time period. The wind direction change rate and direction stability index are calculated. The confidence interval of the prediction results is dynamically corrected to generate wind direction change trend prediction results with time series characteristics. The yaw decision module is used to calculate the predicted deviation angle between the nacelle and the wind direction based on the predicted wind direction change trend and the current nacelle orientation of the wind turbine. At the same time, it sets the dynamic yaw start threshold under different wind direction change trends and wind speed conditions. When the predicted deviation angle is greater than the corresponding dynamic yaw start threshold, the load yaw command is triggered. The load-bearing yaw execution control module is used to control the yaw motor to drive the yaw reduction gearbox, yaw drive gearbox and yaw bearing to work together in the continuous power generation state of the wind turbine after receiving the load-bearing yaw command. This drives the nacelle to rotate in the wind direction. During the rotation, the module monitors the wind turbine's operating parameters in real time, dynamically adjusts the yaw speed and motor output power based on the monitoring data, and synchronously adjusts the wind turbine pitch angle to balance the wind load in the wind turbine plane. The yaw termination module is used to control the yaw motor to stop working, activate the yaw braking device and fix the nacelle position when the deviation angle between the nacelle and the wind direction is less than the set yaw termination threshold, thus completing the yaw adjustment. The parameter optimization module is used to normalize the operational data during this yaw process, construct a yaw operation feature dataset, optimize the dynamic yaw initiation threshold using a clustering analysis algorithm based on this dataset, and perform global optimization of the yaw adjustment parameters using a genetic algorithm. The optimized dynamic yaw initiation threshold and yaw adjustment parameter combination are then fed back into the hybrid prediction model to update the model parameters.

9. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store a computer program, and the processor running the computer program to cause the electronic device to perform the method for optimizing wind turbine power generation based on IPC load yaw strategy as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for optimizing wind turbine power generation based on IPC-based load yaw strategy as described in any one of claims 1-7.

Images