Wind power generation control system and control method

Abstract

The invention relates to the field of wind power generation, and discloses a wind power generation control system and method. The data acquisition module is used for acquiring front three-dimensional wind field information detected by the laser radar, near-field wind speed and direction information measured by the ultrasonic sensor, regional wind regime data issued by the meteorological satellite and operation parameters of the wind generating set. According to the invention, through multi-source information fusion, wind regime accurate perception is realized, power generation stability is improved from the source, through multi-source fusion of the laser radar, the ultrasonic sensor and meteorological satellite data, a three-dimensional wind regime perception system is constructed, wind speed and wind direction changes can be predicted in advance, wind regime measurement errors are reduced, and power generation efficiency is improved. The problem of power generation fluctuation caused by wind energy intermittency is solved from the source, and accurate data support is provided for dynamic adjustment of the pitch angle.

Description

Technical Field

[0001] The present invention relates to the field of wind power generation, and in particular to a wind power generation control system and a control method. Background Art

[0002] As a vital component of clean energy, wind power plays a key role in the global energy transition. The inherent intermittent and fluctuating nature of wind energy poses significant challenges to the stable operation of wind power systems. The magnitude and direction of wind are constantly changing, leading to highly unstable wind energy captured by wind turbines and, in turn, fluctuations in output parameters such as voltage and frequency. In existing wind turbine control systems, the pitch angle adjustment system is one of the core links to ensure stable operation and efficient power generation of the unit. However, traditional pitch angle adjustment systems mostly use fixed logic or simple PID control algorithms, which have significant limitations.

[0003] A search revealed that publication number CN119957420A discloses a wind power generation control system and control method, comprising: a data module for acquiring wind turbine operating data, the operating data including wind speed, wind direction, rotor speed, and wind turbine power; a control module for selecting an initial pitch angle based on the operating data and a pre-set pitch angle selection rule; a first correction module for determining whether a first correction is required to the initial pitch angle after control by the control module; a second correction module for determining whether a second correction is required to the current pitch angle after correction by the first correction module; and a safety module for determining whether an emergency stop control of the wind turbine is required based on the operating data.

[0004] Existing wind power generation control systems are unable to accurately sense changes in wind conditions in real time, and are difficult to dynamically adjust according to the actual power requirements of wind turbines. In strong wind environments, the pitch angle cannot be increased in time, and excess wind energy cannot be effectively controlled, exacerbating the instability of output power. In weak wind conditions, the pitch angle cannot be effectively reduced, resulting in low wind energy conversion efficiency and energy waste. Summary of the Invention

[0005] In order to solve the above-mentioned problem that the existing wind power generation control system cannot perceive the wind condition changes in real time and accurately, and is difficult to dynamically adjust according to the actual power demand of the wind turbine, the present invention is achieved through the following technical solutions.

[0006] The present invention provides a wind power generation control system, comprising: Multi-source information acquisition module, used to collect the three-dimensional wind field information ahead detected by lidar, the near-field wind speed and direction information measured by ultrasonic sensors, the regional wind condition data released by meteorological satellites, and the operating parameters of wind turbines; The intelligent prediction and decision-making module uses a deep learning algorithm to analyze and process the information collected by the multi-source information acquisition module, establish a wind condition prediction model and a pitch angle optimization decision model, and generate pitch angle adjustment instructions based on the predicted wind conditions and the current operating status of the unit; A distributed collaborative control module includes a main controller and multiple sub-controllers. The main controller receives the pitch angle adjustment instructions generated by the intelligent prediction and decision module and coordinates the work of each sub-controller. Each sub-controller executes the pitch angle adjustment instructions to control the corresponding actuator; The adaptive variable structure control module is used to automatically adjust the structure and parameters of the control algorithm according to different wind conditions and unit operating status. Under normal wind conditions, it adopts a pitch angle adjustment strategy based on model predictive control and switches to a robust control strategy under extreme wind conditions.

[0007] Preferably, the multi-source information acquisition module includes: The laser radar unit consists of a laser transmitter, a receiver, a scanning mechanism, and a data processing circuit. The laser transmitter emits a laser beam to detect the wind field ahead, the receiver receives the reflected signal, the scanning mechanism realizes three-dimensional space scanning, and the data processing circuit converts the signal into information about wind speed, wind direction, and turbulence intensity. The ultrasonic sensor unit includes an ultrasonic transmitting probe, a receiving probe, a timing circuit, and a signal processing chip. The ultrasonic transmitting probe emits ultrasonic waves, the receiving probe receives the reflected waves, the timing circuit calculates the propagation time, and the signal processing chip calculates the wind speed and direction based on this time. The meteorological satellite data receiving unit consists of a satellite communication antenna, a signal demodulation module, and a data decoding module. The satellite communication antenna receives satellite signals, the signal demodulation module restores the data signal, and the data decoding module parses the regional wind condition data. The unit operation parameter sensor unit includes a speed sensor, a torque sensor and a power sensor, which are used to collect the speed, torque and output power parameters of the wind turbine generator set respectively.

[0008] Preferably, the intelligent prediction and decision-making module includes: The data preprocessing unit cleans and normalizes the data transmitted by the multi-source information acquisition module to provide high-quality data for subsequent modeling; The deep learning model training unit, based on the TensorFlow or PyTorch framework, uses historical wind data and turbine operation data to train and optimize the LSTM wind prediction model and the CNN combined with reinforcement learning pitch angle optimization decision model. The prediction and decision execution unit calls the trained model in real time to predict the current and future wind conditions, and generates the optimal pitch angle adjustment instructions based on the unit's operating status.

[0009] Preferably, the main controller includes a processor, memory, storage chip and communication interface circuit, and is responsible for receiving instructions from the intelligent prediction and decision-making module and coordinating the work of each sub-controller.

[0010] Preferably, the sub-controller includes: The pitch angle controller is connected to the drive circuit and the feedback circuit, executes the pitch angle adjustment command and feeds back the adjustment status; The converter controller adjusts the converter's operating parameters through the control signal interface to achieve power conversion control; The yaw controller, combined with the motor drive circuit and position detection circuit, controls the yaw system to adjust the windward direction of the unit; The communication network unit includes a network interface chip and a network transmission cable to realize data communication between the main controller and the sub-controller.

[0011] Preferably, the adaptive variable structure control module includes: The model building unit builds a dynamic mathematical model based on the physical characteristics and operating principles of the wind turbine generator set to calculate the control parameters under different working conditions; The status monitoring unit collects wind conditions and unit operating status data in real time, monitors system operation, and determines whether the control algorithm needs to be adjusted; The algorithm switching and parameter optimization unit automatically switches the control algorithm through adaptive switching logic according to the status monitoring results, and uses genetic algorithm or particle swarm optimization algorithm to perform online optimization of the control parameters.

[0012] Preferably, the actuator includes: The pitch angle adjustment mechanism unit consists of an electric servo motor, a hydraulic pump, a hydraulic cylinder, a transmission mechanism, and a position sensor. The electric servo motor realizes position control, the hydraulic pump and hydraulic cylinder provide high torque drive, the transmission mechanism transmits power to the blades, and the position sensor provides feedback on the actual position of the pitch angle. The full-power converter unit, including the rectifier circuit, DC link, inverter circuit and control circuit, realizes the frequency, voltage and phase regulation of the output power of the wind turbine generator set; The yaw system unit consists of a yaw motor, a reducer, a yaw bearing and a wind direction sensor, and is used to adjust the windward direction of the unit according to the wind direction sensor signal.

[0013] Preferably, the wind condition prediction model uses a long short-term memory network to learn historical wind condition data and predict the wind speed and wind direction change trends within the next 10-30 minutes; The pitch angle optimization decision model uses a convolutional neural network combined with a reinforcement learning algorithm to calculate the optimal pitch angle adjustment scheme.

[0014] Preferably, it also includes a multi-source information fusion processing module, which uses the Kalman filter algorithm to filter the real-time wind condition data collected by lidar, ultrasonic sensors, etc. to remove noise interference, and uses the DS evidence theory to fuse different types of sensor data and meteorological satellite data to construct a unified wind condition information model.

[0015] The present invention also provides a wind power generation control method, comprising the following steps: The system uses lidar to detect three-dimensional wind field information within a range of 300-500 meters ahead, ultrasonic sensors to measure near-field wind speed and direction information, and meteorological satellite data receiving modules to obtain regional wind condition data and collect wind turbine operating parameters. The Kalman filter algorithm is used to filter the collected real-time wind data, and the DS evidence theory is used to fuse different types of sensor data and meteorological satellite data to build a unified wind information model. The fused historical wind data is learned using a long-short-term memory network to predict the wind speed and direction trends within the next 10-30 minutes. A convolutional neural network combined with a reinforcement learning algorithm is used to calculate and generate the optimal pitch angle adjustment command based on the predicted wind conditions and the current operating status of the turbine. The main controller receives pitch angle adjustment instructions and coordinates the work of multiple sub-controllers. Each sub-controller executes the pitch angle adjustment instructions to control the corresponding actuator. The main controller and each sub-controller communicate at a rate of no less than 100Mbps through a communication protocol stack based on the Modbus TCP / IP protocol. According to different wind conditions and unit operating status, a dynamic mathematical model of the wind turbine is established, and the control parameters and structural adjustment strategy of the system are calculated in real time. Through adaptive switching logic, a pitch angle adjustment strategy based on model predictive control is adopted under normal wind conditions, and a robust control strategy is switched to under extreme wind conditions, and the control parameters are adjusted using genetic algorithms or particle swarm optimization algorithms.

[0016] The present invention provides a wind power generation control system and control method. Compared with the existing technology, it has the following advantages: Multi-source information fusion enables accurate perception of wind conditions, improving power generation stability from the source. Through the multi-source fusion of lidar, ultrasonic sensors and meteorological satellite data, a three-dimensional wind condition perception system is constructed, which can predict changes in wind speed and direction in advance, reduce wind condition measurement errors, and solve the power generation fluctuation problem caused by intermittent wind energy from the source, providing accurate data support for dynamic adjustment of pitch angle.

[0017] An LSTM network is introduced to predict wind trends, and a pitch angle optimization model is constructed using CNN and reinforcement learning, replacing traditional PID or fixed threshold control. This algorithm autonomously learns from historical data the characteristics of different wind fields, increasing the pitch angle in advance of strong winds to limit wind energy absorption and dynamically reducing the pitch angle in weak winds to improve capture efficiency.

[0018] The distributed collaborative architecture improves the system's response speed and fault tolerance. It adopts a distributed architecture of the main controller and sub-controllers to achieve real-time data interaction. When a module fails, the main controller can automatically switch to the redundant control path, and other sub-controllers take over some functions. The system's fault tolerance is improved, and the collaborative adjustment accuracy of each module is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is the overall architecture diagram of the system proposed in this invention.

[0020] Figure 2 This is a block diagram of the multi-source information acquisition module proposed in the present invention.

[0021] Figure 3 This is a block diagram of the intelligent prediction and decision-making module proposed in this invention.

[0022] Figure 4 This is a block diagram of the distributed collaborative control module proposed in this invention. DETAILED DESCRIPTION

[0023] The present invention is further described below with reference to specific examples. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of protection of the present invention.

[0024] The following describes the embodiments of the present invention through specific examples. Those skilled in the art will readily understand the other advantages and benefits of the present invention from the disclosure herein. The present invention may also be implemented or applied through various other specific embodiments, and the details in this specification may be modified or altered based on different viewpoints and applications without departing from the spirit of the present invention.

[0025] Example 1: Reference Figures 1-4 , a wind power generation control system, comprising: Multi-source information acquisition module, used to collect the three-dimensional wind field information ahead detected by lidar, the near-field wind speed and direction information measured by ultrasonic sensors, the regional wind condition data released by meteorological satellites, and the operating parameters of wind turbines; In the multi-source information acquisition module, the lidar is installed on the top of the wind turbine tower to detect wind speed, wind direction, and turbulence intensity information within a range of 300-500 meters in front. The detection frequency is not less than 10Hz, and the measurement accuracy of wind speed error is ±0.5m / s, and wind direction error is ±1°; the ultrasonic sensor is arranged on the surface of the cabin, with a sampling frequency of 50Hz, a wind speed measurement range of 0-60m / s, an accuracy of ±0.1m / s, and a wind direction measurement range of 0-360°, with an accuracy of ±1°; the meteorological satellite data receiving module receives data through a satellite communication antenna, and the update frequency is once every 15 minutes.

[0026] The intelligent prediction and decision-making module uses a deep learning algorithm to analyze and process the information collected by the multi-source information acquisition module, establish a wind condition prediction model and a pitch angle optimization decision model, and generate pitch angle adjustment instructions based on the predicted wind conditions and the current operating status of the unit; In the intelligent prediction and decision-making module, the wind condition prediction model uses a long short-term memory network to learn historical wind condition data and predict the wind speed and wind direction change trends in the next 10-30 minutes; the pitch angle optimization decision model uses a convolutional neural network combined with a reinforcement learning algorithm to calculate the optimal pitch angle adjustment plan.

[0027] Distributed collaborative control module, which includes a main controller and multiple sub-controllers. The main controller receives the pitch angle adjustment command generated by the intelligent prediction and decision module and coordinates the work of each sub-controller. Each sub-controller executes the pitch angle adjustment command to control the corresponding actuator; In the distributed collaborative control module, the main controller uses the ARMCortex-A72 series processor, and the sub-controller uses the STM32F7 series microcontroller. The main controller and each sub-controller communicate through a communication protocol stack based on the ModbusTCP / IP protocol, and the communication rate is not less than 100Mbps.

[0028] The adaptive variable structure control module is used to automatically adjust the structure and parameters of the control algorithm according to different wind conditions and unit operating status. Under normal wind conditions, it adopts a pitch angle adjustment strategy based on model predictive control and switches to a robust control strategy under extreme wind conditions.

[0029] The adaptive variable structure control module establishes a dynamic mathematical model of the wind turbine generator set, calculates the system's control parameters and structural adjustment strategy in real time, and automatically switches the control algorithm through adaptive switching logic when wind conditions change or abnormal unit operating status is detected, and uses genetic algorithm or particle swarm optimization algorithm to adjust the control parameters.

[0030] In this embodiment, the multi-source information acquisition module collects wind conditions and unit operation data, which are pre-processed by the multi-source information fusion processing module and then input into the intelligent prediction and decision-making module. The intelligent prediction and decision-making module uses a deep learning algorithm to analyze and process the data, predict wind conditions and generate pitch angle adjustment instructions. After receiving the instructions, the main controller of the distributed collaborative control module coordinates each sub-controller to control the actuator to perform the corresponding action. The operating status of the actuator and the new operating parameters of the unit are fed back to the multi-source information acquisition module through sensors, and enter the data processing and decision-making process again. The adaptive variable structure control module adjusts the control algorithm and parameters in real time according to the system operating status to form a closed-loop control. The entire solution forms a complete closed-loop control system from data acquisition, analysis and decision-making, instruction execution to feedback optimization, ensuring that the system can continuously optimize operation according to wind conditions and unit status.

[0031] This solution integrates multi-source information such as lidar, ultrasonic sensors, and meteorological satellite data to achieve comprehensive, accurate, and forward-looking perception of wind conditions, with a wider detection range and richer data dimensions; it introduces a deep learning algorithm that can autonomously learn historical data to achieve accurate prediction of wind conditions and forward-looking dynamic adjustment of the pitch angle; it adopts a distributed collaborative control architecture, in which each module operates independently and works together, improving the system's response speed and fault tolerance, and the failure of a module will not affect the overall operation; the control algorithm is automatically switched according to the wind conditions and unit status, and the optimization algorithm is used to dynamically adjust the parameters to ensure the unit's efficient and safe operation under various operating conditions. Through multi-module collaboration and feedback mechanism, a complete closed loop from data acquisition to optimization control is formed to continuously improve system performance.

[0032] The multi-source information acquisition module includes: The laser radar unit consists of a laser transmitter, a receiver, a scanning mechanism, and a data processing circuit. The laser transmitter emits a laser beam to detect the wind field ahead, the receiver receives the reflected signal, the scanning mechanism realizes three-dimensional space scanning, and the data processing circuit converts the signal into information about wind speed, wind direction, and turbulence intensity. The ultrasonic sensor unit includes an ultrasonic transmitting probe, a receiving probe, a timing circuit, and a signal processing chip. The ultrasonic transmitting probe emits ultrasonic waves, the receiving probe receives the reflected waves, the timing circuit calculates the propagation time, and the signal processing chip calculates the wind speed and direction based on this time. The meteorological satellite data receiving unit consists of a satellite communication antenna, a signal demodulation module, and a data decoding module. The satellite communication antenna receives satellite signals, the signal demodulation module restores the data signal, and the data decoding module parses the regional wind condition data. The unit operation parameter sensor unit includes a speed sensor, a torque sensor and a power sensor, which are used to collect the speed, torque and output power parameters of the wind turbine generator set respectively.

[0033] Intelligent prediction and decision-making modules include: The data preprocessing unit cleans and normalizes the data transmitted by the multi-source information acquisition module to provide high-quality data for subsequent modeling; The deep learning model training unit, based on the TensorFlow or PyTorch framework, uses historical wind data and turbine operation data to train and optimize the LSTM wind prediction model and the CNN combined with reinforcement learning pitch angle optimization decision model. The prediction and decision execution unit calls the trained model in real time to predict the current and future wind conditions, and generates the optimal pitch angle adjustment instructions based on the unit's operating status.

[0034] The main controller includes a processor, memory, storage chip and communication interface circuit. It is responsible for receiving instructions from the intelligent prediction and decision-making module and coordinating the work of each sub-controller.

[0035] The sub-controllers include: The pitch angle controller, based on the STM32F7 series microcontroller, connects the drive circuit and feedback circuit, executes the pitch angle adjustment command and feedbacks the adjustment status; The converter controller also uses the STM32F7 single-chip microcomputer to adjust the converter's operating parameters through the control signal interface to achieve power conversion control; The yaw controller, with a single-chip microcomputer as its core, combines the motor drive circuit and the position detection circuit to control the yaw system to adjust the windward direction of the unit; The communication network unit is built based on the Modbus TCP / IP protocol stack and includes a network interface chip and a network transmission cable to realize data communication between the main controller and the sub-controller.

[0036] The adaptive variable structure control module includes: The model building unit builds a dynamic mathematical model based on the physical characteristics and operating principles of the wind turbine generator set to calculate the control parameters under different working conditions; The status monitoring unit collects wind conditions and unit operating status data in real time, monitors system operation, and determines whether the control algorithm needs to be adjusted; The algorithm switching and parameter optimization unit automatically switches the control algorithm through adaptive switching logic according to the status monitoring results, and uses genetic algorithm or particle swarm optimization algorithm to perform online optimization of the control parameters.

[0037] The implementing agencies include: The pitch angle adjustment mechanism unit consists of an electric servo motor, a hydraulic pump, a hydraulic cylinder, a transmission mechanism, and a position sensor. The electric servo motor realizes position control, the hydraulic pump and hydraulic cylinder provide high torque drive, the transmission mechanism transmits power to the blades, and the position sensor provides feedback on the actual position of the pitch angle. The full-power converter unit, including the rectifier circuit, DC link, inverter circuit and control circuit, realizes the frequency, voltage and phase regulation of the output power of the wind turbine generator set; The yaw system unit consists of a yaw motor, a reducer, a yaw bearing and a wind direction sensor, and is used to adjust the windward direction of the unit according to the wind direction sensor signal.

[0038] The wind forecast model uses a long-short-term memory network to learn historical wind data and predict the wind speed and direction trends within the next 10-30 minutes. The pitch angle optimization decision model uses a convolutional neural network combined with a reinforcement learning algorithm to calculate the optimal pitch angle adjustment scheme.

[0039] It also includes a multi-source information fusion processing module, which uses the Kalman filter algorithm to filter the real-time wind data collected by lidar, ultrasonic sensors, etc. to remove noise interference, and uses the DS evidence theory to fuse different types of sensor data and meteorological satellite data to build a unified wind information model.

[0040] The multi-source information acquisition module transmits the raw data collected by the lidar, ultrasonic sensor, meteorological satellite data receiving unit and unit operation parameter sensor unit to the multi-source information fusion processing module through a data transmission cable or bus (such as CAN bus, RS-485 bus).

[0041] The wind condition information model after data filtering and fusion processing is input into the data preprocessing unit of the intelligent prediction and decision-making module through the Ethernet interface.

[0042] The pitch angle adjustment command generated by the intelligent prediction and decision module is transmitted to the main controller of the distributed collaborative control module through a communication network based on the Modbus TCP / IP protocol.

[0043] After receiving the instruction, the main controller connects each sub-controller with the dedicated control cable or interface of the corresponding unit of the actuator and sends the control signal.

[0044] The position sensor, wind direction sensor, etc. in the actuator feeds back the operating status of the actuator and the new operating parameters of the unit to the unit operating parameter sensor unit in the multi-source information acquisition module.

[0045] The state monitoring unit of the adaptive variable structure control module collects wind condition data from the multi-source information acquisition module and unit operation status data fed back by the actuator in real time for algorithm switching and parameter optimization judgment.

[0046] In the multi-source information acquisition module, the laser transmitter, receiver, and scanning mechanism are connected to the data processing circuit through internal electrical circuits, and the data processed by the data processing circuit is connected to the external module through the data output interface; the ultrasonic transmitting probe and receiving probe are connected to the timing circuit, and the timing circuit transmits the time data to the signal processing chip, and the output end of the signal processing chip is connected to the external data transmission line; after the satellite communication antenna receives the signal, it is processed by the signal demodulation module and then transmitted to the data decoding module. The output interface of the data decoding module is connected to the external data transmission line; the speed sensor, torque sensor, and power sensor transmit the collected analog signal to the signal conditioning circuit through the signal line, and after A / D conversion, it is output to the external data transmission line in the form of a digital signal.

[0047] In the intelligent prediction and decision-making module, the input interface is connected to the output end of the multi-source information fusion processing module. After cleaning and normalizing the received data, it is transmitted to the deep learning model training unit and the prediction and decision execution unit through the internal data bus; historical data is obtained through the data interface for model training, and the trained model parameters are stored in the internal storage unit and shared with the prediction and decision execution unit through the data bus; real-time data is obtained from the data preprocessing unit, and the trained model is called to predict wind conditions and generate pitch angle adjustment instructions, and the instructions are transmitted to the distributed collaborative control module through the output interface.

[0048] In the distributed collaborative control module, the module is connected to the intelligent prediction and decision-making module through the Modbus TCP / IP protocol communication network interface to receive instructions, and exchanges data and distributes instructions with each sub-controller through the internal data bus. At the same time, it is connected to external devices (such as the monitoring system) through the communication interface. The input interface receives instructions from the main controller and controls the electric servo motor and hydraulic system of the pitch angle adjustment mechanism unit through the drive circuit. The actual pitch angle position signal fed back by the position sensor is transmitted back to the pitch angle controller through the feedback circuit. The converter controller communicates with the main controller to receive instructions, adjusts the operating parameters of the rectifier circuit, inverter circuit, and other working parameters of the full-power converter unit through the control signal interface, and receives the converter operation status feedback signal. After receiving the instructions from the main controller, the yaw controller controls the yaw motor through the motor drive circuit, and the position detection circuit on the yaw bearing feeds back the windward direction information of the unit to the yaw controller. The network interface chip is connected to the communication interface of the main controller and each sub-controller, and a communication network is established through network transmission cables to achieve high-speed and reliable data transmission.

[0049] In the adaptive variable structure control module, the physical parameters and operating principle-related data of the wind turbine generator set are obtained through the data interface to establish a dynamic mathematical model. The model parameters are stored in the internal storage unit and can be shared with the algorithm switching and parameter optimization unit through the data bus. The wind condition data of the multi-source information acquisition module and the unit operation status data fed back by the actuator are collected in real time through the data acquisition interface. After data processing, they are transmitted to the algorithm switching and parameter optimization unit. According to the data from the status monitoring unit, the control algorithm is switched through the adaptive switching logic, and the control parameters are optimized online using the genetic algorithm or the particle swarm optimization algorithm. The optimized parameters are transmitted to the sub-controllers of the distributed collaborative control module through the data bus.

[0050] In the multi-source information fusion processing module, the input interface of the data filtering unit is connected to the output end of the multi-source information acquisition module, and Kalman filtering is performed on the received real-time wind condition data. The processed data is transmitted to the data fusion unit through the internal data bus; the data fusion unit receives the data from the data filtering unit and the meteorological satellite data, uses the DS evidence theory to perform fusion processing, constructs a unified wind condition information model, and transmits it to the intelligent prediction and decision-making module through the output interface.

[0051] In the actuator, the electric servo motor, hydraulic pump, and hydraulic cylinder are connected to the blades through a transmission mechanism. The electric servo motor and hydraulic system are controlled by the drive circuit of the pitch angle controller. The position sensor detects the pitch angle in real time and feeds back the signal to the pitch angle controller; the rectifier circuit, DC link, inverter circuit, and control circuit are connected in sequence through electrical lines. The control circuit receives instructions from the converter controller to adjust the working parameters and realize electrical energy conversion; the yaw motor is connected to the yaw bearing through a reducer, and the wind direction sensor detects the wind direction signal and transmits it to the yaw controller. The yaw controller controls the yaw motor according to the instructions to adjust the windward direction of the unit.

[0052] Embodiment 2: A wind power generation control method, comprising the following steps: The system uses lidar to detect three-dimensional wind field information within a range of 300-500 meters ahead, ultrasonic sensors to measure near-field wind speed and direction information, and meteorological satellite data receiving modules to obtain regional wind condition data and collect wind turbine operating parameters. The Kalman filter algorithm is used to filter the collected real-time wind data, and the DS evidence theory is used to fuse different types of sensor data and meteorological satellite data to build a unified wind information model. The fused historical wind data is learned using a long-short-term memory network to predict the wind speed and direction trends within the next 10-30 minutes. A convolutional neural network combined with a reinforcement learning algorithm is used to calculate and generate the optimal pitch angle adjustment command based on the predicted wind conditions and the current operating status of the turbine. The main controller receives pitch angle adjustment instructions and coordinates the work of multiple sub-controllers. Each sub-controller executes the pitch angle adjustment instructions to control the corresponding actuator. The main controller and each sub-controller communicate at a rate of no less than 100Mbps through a communication protocol stack based on the Modbus TCP / IP protocol. According to different wind conditions and unit operating status, a dynamic mathematical model of the wind turbine is established, and the control parameters and structural adjustment strategy of the system are calculated in real time. Through adaptive switching logic, a pitch angle adjustment strategy based on model predictive control is adopted under normal wind conditions, and a robust control strategy is switched to under extreme wind conditions, and the control parameters are adjusted using genetic algorithms or particle swarm optimization algorithms.

[0053] In summary, compared with the existing technology, it has the following beneficial effects: Multi-source information fusion enables accurate perception of wind conditions, improving power generation stability from the source. Through the multi-source fusion of lidar, ultrasonic sensors and meteorological satellite data, a three-dimensional wind condition perception system is constructed, which can predict changes in wind speed and direction in advance, reduce wind condition measurement errors, and solve the power generation fluctuation problem caused by intermittent wind energy from the source, providing accurate data support for dynamic adjustment of pitch angle.

[0054] An LSTM network is introduced to predict wind trends, and a pitch angle optimization model is constructed using CNN and reinforcement learning, replacing traditional PID or fixed threshold control. This algorithm autonomously learns from historical data the characteristics of different wind fields, increasing the pitch angle in advance of strong winds to limit wind energy absorption and dynamically reducing the pitch angle in weak winds to improve capture efficiency.

[0055] The distributed collaborative architecture improves the system's response speed and fault tolerance. It adopts a distributed architecture of the main controller and sub-controllers to achieve real-time data interaction. When a module fails, the main controller can automatically switch to the redundant control path, and other sub-controllers take over some functions. The system's fault tolerance is improved, and the collaborative adjustment accuracy of each module is improved.

[0056] Thus, although the invention has been described herein with reference to specific embodiments thereof, freedom of modification, various changes and substitutions are contemplated within the foregoing disclosure, and it should be understood that in some cases, some features of the invention will be employed without the corresponding use of other features without departing from the scope and spirit of the claimed invention. Thus, many modifications may be made to adapt a particular environment or material to the true scope and spirit of the invention. The invention is not intended to be limited to the specific terminology used in the claims below and / or to the specific embodiments disclosed as the best mode contemplated for carrying out the invention, but the invention is intended to include any and all embodiments and equivalents falling within the scope of the appended claims. Thus, the scope of the invention will be determined solely by the appended claims.

Claims

1. A wind power generation control system, characterized in that: include: Multi-source information acquisition module, used to collect the three-dimensional wind field information ahead detected by lidar, the near-field wind speed and direction information measured by ultrasonic sensors, the regional wind condition data released by meteorological satellites, and the operating parameters of wind turbines; The intelligent prediction and decision-making module uses a deep learning algorithm to analyze and process the information collected by the multi-source information acquisition module, establish a wind condition prediction model and a pitch angle optimization decision model, and generate pitch angle adjustment instructions based on the predicted wind conditions and the current operating status of the unit; A distributed collaborative control module includes a main controller and multiple sub-controllers. The main controller receives the pitch angle adjustment instructions generated by the intelligent prediction and decision module and coordinates the work of each sub-controller. Each sub-controller executes the pitch angle adjustment instructions to control the corresponding actuator; The adaptive variable structure control module is used to automatically adjust the structure and parameters of the control algorithm according to different wind conditions and unit operating status. Under normal wind conditions, it adopts a pitch angle adjustment strategy based on model predictive control and switches to a robust control strategy under extreme wind conditions.

2. The wind power generation control system according to claim 1, characterized in that: The multi-source information acquisition module includes: The laser radar unit consists of a laser transmitter, a receiver, a scanning mechanism, and a data processing circuit. The laser transmitter emits a laser beam to detect the wind field ahead, the receiver receives the reflected signal, the scanning mechanism realizes three-dimensional space scanning, and the data processing circuit converts the signal into information about wind speed, wind direction, and turbulence intensity. The ultrasonic sensor unit includes an ultrasonic transmitting probe, a receiving probe, a timing circuit, and a signal processing chip. The ultrasonic transmitting probe emits ultrasonic waves, the receiving probe receives the reflected waves, the timing circuit calculates the propagation time, and the signal processing chip calculates the wind speed and direction based on this time. The meteorological satellite data receiving unit consists of a satellite communication antenna, a signal demodulation module, and a data decoding module. The satellite communication antenna receives satellite signals, the signal demodulation module restores the data signal, and the data decoding module parses the regional wind condition data. The unit operation parameter sensor unit includes a speed sensor, a torque sensor and a power sensor, which are used to collect the speed, torque and output power parameters of the wind turbine generator set respectively.

3. The wind power generation control system according to claim 2, characterized in that: The intelligent prediction and decision-making module includes: The data preprocessing unit cleans and normalizes the data transmitted by the multi-source information acquisition module to provide high-quality data for subsequent modeling; The deep learning model training unit, based on the TensorFlow or PyTorch framework, uses historical wind data and turbine operation data to train and optimize the LSTM wind prediction model and the CNN combined with reinforcement learning pitch angle optimization decision model. The prediction and decision execution unit calls the trained model in real time to predict the current and future wind conditions, and generates the optimal pitch angle adjustment instructions based on the unit's operating status.

4. The wind power generation control system according to claim 1, characterized in that: The main controller includes a processor, memory, storage chip and communication interface circuit, and is responsible for receiving instructions from the intelligent prediction and decision-making module and coordinating the work of each sub-controller.

5. The wind power generation control system according to claim 4, characterized in that: The sub-controller includes: The pitch angle controller is connected to the drive circuit and the feedback circuit, executes the pitch angle adjustment command and feeds back the adjustment status; The converter controller adjusts the converter's operating parameters through the control signal interface to achieve power conversion control; The yaw controller, combined with the motor drive circuit and position detection circuit, controls the yaw system to adjust the windward direction of the unit; The communication network unit includes a network interface chip and a network transmission cable to realize data communication between the main controller and the sub-controller.

6. The wind power generation control system according to claim 1, characterized in that: The adaptive variable structure control module includes: The model building unit builds a dynamic mathematical model based on the physical characteristics and operating principles of the wind turbine generator set to calculate the control parameters under different working conditions; The status monitoring unit collects wind conditions and unit operating status data in real time, monitors system operation, and determines whether the control algorithm needs to be adjusted; The algorithm switching and parameter optimization unit automatically switches the control algorithm through adaptive switching logic according to the status monitoring results, and uses genetic algorithm or particle swarm optimization algorithm to perform online optimization of the control parameters.

7. The wind power generation control system according to claim 1, characterized in that: The executive mechanism comprises: The pitch angle adjustment mechanism unit consists of an electric servo motor, a hydraulic pump, a hydraulic cylinder, a transmission mechanism, and a position sensor. The electric servo motor realizes position control, the hydraulic pump and hydraulic cylinder provide high torque drive, the transmission mechanism transmits power to the blades, and the position sensor provides feedback on the actual position of the pitch angle. The full-power converter unit, including the rectifier circuit, DC link, inverter circuit and control circuit, realizes the frequency, voltage and phase regulation of the output power of the wind turbine generator set; The yaw system unit consists of a yaw motor, a reducer, a yaw bearing and a wind direction sensor, and is used to adjust the windward direction of the unit according to the wind direction sensor signal.

8. The wind power generation control system according to claim 1, characterized in that: The wind condition prediction model uses a long short-term memory network to learn historical wind condition data and predict the wind speed and wind direction change trends within the next 10-30 minutes; The pitch angle optimization decision model uses a convolutional neural network combined with a reinforcement learning algorithm to calculate the optimal pitch angle adjustment scheme.

9. The wind power generation control system according to claim 1, characterized in that: It also includes a multi-source information fusion processing module, which uses the Kalman filter algorithm to filter the real-time wind data collected by lidar, ultrasonic sensors, etc. to remove noise interference, and uses the DS evidence theory to fuse different types of sensor data and meteorological satellite data to build a unified wind information model.

10. A wind power generation control method, characterized in that: The following steps are involved: The system uses lidar to detect three-dimensional wind field information within a range of 300-500 meters ahead, ultrasonic sensors to measure near-field wind speed and direction information, and meteorological satellite data receiving modules to obtain regional wind condition data and collect wind turbine operating parameters. The Kalman filter algorithm is used to filter the collected real-time wind data, and the DS evidence theory is used to fuse different types of sensor data and meteorological satellite data to build a unified wind information model. The fused historical wind data is learned using a long-short-term memory network to predict the wind speed and direction trends within the next 10-30 minutes. A convolutional neural network combined with a reinforcement learning algorithm is used to calculate and generate the optimal pitch angle adjustment command based on the predicted wind conditions and the current operating status of the turbine. The main controller receives pitch angle adjustment instructions and coordinates the work of multiple sub-controllers. Each sub-controller executes the pitch angle adjustment instructions to control the corresponding actuator. The main controller and each sub-controller communicate at a rate of no less than 100Mbps through a communication protocol stack based on the Modbus TCP / IP protocol. According to different wind conditions and unit operating status, a dynamic mathematical model of the wind turbine is established, and the control parameters and structural adjustment strategy of the system are calculated in real time. Through adaptive switching logic, a pitch angle adjustment strategy based on model predictive control is adopted under normal wind conditions, and a robust control strategy is switched to under extreme wind conditions, and the control parameters are adjusted using genetic algorithms or particle swarm optimization algorithms.

Images