Research On Pitch Optimal Control Of Wind Turbines Based On Neural Network

Wind power generation is very environmentally friendly and rich in resources,vigorously developing the wind power industry can alleviate the adverse effects brought by the energy crisis and ensure the healthy development of human society.The pitch control technology is one of the core technologies in the wind power industry,the continuous development of this technology is conducive to the prosperity of the wind power industry.Therefore,the pitch control are the focus in the paper,the main research contents are as follows:Firstly,the basic principles of pitch control and the characteristics of wind turbine loads are described.Wind turbine output power is fluctuated easily due to the randomness of wind speed and the nonlinearity and time lag of wind turbines.For small-scale wind turbines,aiming at solving the difficulty to achieve good control effect for the collection pitch control based on the traditional PID or PI control,a pitch control strategy based on proportional controller and extreme learning machine(ELM)is proposed.The steady-state output of the PI controller of wind turbines at each wind speed is learned by ELM.The trained ELM and a proportional controller are then adopted to control the wind turbine pitches.The simulation results show that the proposed strategy has high control accuracy,short response time,and the control effect in stabilizing the wind turbine output power is better than that of traditional PI controllers.Secondly,large unbalanced loads are made during operation for large-scale wind turbines due to wind shear and tower shadow effects.Aiming at solving this problem,a wind turbine provided by FAST software is used as the research object and an individual pitch control strategy based on azimuth feedback is used to simulate.Compared with the collection pitch control,the unbalanced loads of wind turbines are reduced to a certain extent.Another individual pitch control strategy based on load feedback is studied,the simulation results show that the proposed method has better control effect.Finally,in order to improve the performance of the individual pitch control strategy based on load feedback,an individual pitch control strategy based on BP neural network and PID is proposed.BP neural network is applied to the individual pitch load PID control loop and its self-learning and weighting coefficient adjustment are performed according to the operating state of wind turbines to realize PID parameter on-line tuning.The simulation results show that the proposed control strategy not only ensures the stability of the wind turbine output power,but also reduces the unbalanced loads more effectively than that of the traditional individual pitch control strategy based on load feedback,thus proving the suitability for the individual pitch control of large-scale wind turbines.