改进单神经元在风力机变桨控制中的研究

Research on Improving Single Neuron in Wind Turbine Pitch Control

针对风力发电机组变桨控制非线性、时变性的问题和传统仿真试验简单数学传递函数模型仿真,采用改进单神经元PID控制器作为变桨控制器,使变桨距控制既具备神经网络控制较强的适应能力,又拥有PID控制较高的精度;并且通过Simulink的simSimmechanics和SimHydraulics仿真,考虑到桨叶的机械特性,以及变桨执行机构的运动特性;最后将改进单神经元PID控制器加入到变桨执行机构的仿真模型中。结果表明,变桨距控制采用改进单神经元PID控制方式的机组,控制响应更加及时,且在变桨控制中调控平稳,对传动系统及变桨执行机构损耗小。

Aiming at the nonlinear and time-varying problems of wind turbine pitch control and the simple mathematical transfer function model simulation of traditional real experiment,the improved single neuron PID controller was used as the pitch controller,so that the pitch control has both neural network control,strong adaptability,and high precision of PID control;and through Simulink’s simSimmechanics and simHydraulics simulation,considering the mechanical characteristics of the blade,the motion characteristics of the pitch actuator.Finally,the improved single neuron PID controller is added to the simulation model of the pitch actuator.The results show that the pitch control uses a unit with improved single neuron PID control mode,the control response is more timely,and the control is stable in the pitch control,and the loss of the transmission system and the pitch actuator is small.