基于数据驱动的风力发电系统一步超前预测转速控制

摘要

In view of characteristics of rotor speed control system of wind power system,the paper presented an one-step ahead predictive control method based on data-driven.Firstly,the method uses dynamic linearization technique and adopts pseudo-projection algorithm to estimate online partial derivative,and then uses the minimizing weighted performance indicators of one-step ahead to get generalized minimum variance predictive control algorithm.In order to achieve better tracking results,an adaptive item was induced to enhance output tracking capability,so make the system have stronger output response tracking.The simulation results show that the method is effective.%针对风力发电系统转速控制系统的特点,提出了一种基于数据驱动的一步超前预测控制方法.该方法先利用动态线性化技术,采用投影算法在线估计伪偏导数,然后通过最小化一步超前加权性能指标获得广义最小方差预测控制算法;为了取得更好的跟踪效果,引入一个自适应项用于增强输出跟踪能力,使得系统输出响应具有更强的跟踪能力.仿真结果验证了该方法的有效性.