风力机变桨系统单神经元自适应PID控制

摘要

在分析风力发电系统能量转换理论的基础上,将单神经元自适应算法与常规PID算法结合,通过单神经元自学习,对PID控制器的系数进行在线调整,提出一种单神经元自适应PID调节变桨角度的控制方法,建立了风力机桨距角控制数学模型.风速超过额定值后,利用单神经元自适应PID控制器进行变桨控制,将控制结果与采用常规PID控制器的控制性能进行了对比.结果表明:单神经元自适应PID控制器可以将输出功率恒定在额定值,能够对变桨系统进行有效控制,达到预期的目标.%Single neuron adaptive algorithm and conventional PID algorithm were combined based on energy transform theory for wind turbine systems. A single neuron adaptive PID pitch angle control method was proposed for online regulation of PID controller parameters via its self-learning capability, while a mathematical model for the pitch angle control created. The pitch angle was then regulated with the proposed PID controller under conditions the wind velocity being more than the rated value, after which a comparison of control performance made between the newly proposed and conventional controllers. Results show that via the single neuron adaptive PID controller, the output power can be maintained at a rated value,through which effective control of pitch angle can be obtained, and therefore expected control purposes can be realized.