基于RBF神经网络的直驱风力发电系统反推控制  

Backstepping Control of Direct-driven Wind Power Generation System Based on RBF Neural Network

作　　者：周尔卓 沈艳霞[1] ZHOU Erzhuo;SHEN Yanxia(School of Internet of Things Engineering,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学物联网工程学院,无锡214122

出　　处：《微特电机》2022年第6期41-45,共5页Small & Special Electrical Machines

摘　　要：考虑风力发电系统的高度非线性及不确定性,提出永磁直驱式风力发电系统反推控制策略,实现最大风能跟踪。利用RBF神经网络的学习和估计能力,设计未知扰动观测器,在逼近系统未知部分的基础上优化控制器设计。仿真结果表明,所用方法提高了系统转速跟踪能力,提高了风能捕获效率。Considering the high nonlinearity and uncertainty of wind power generation system,a reverse thrust control strategy for permanent magnet direct drive wind power generation system was proposed to achieve maximum wind energy tracking.Using the learning and estimation ability of the RBF neural network,an unknown disturbance observer was designed,and the controller design was optimized on the basis of approximating the unknown part of the system.The simulation results show that the method used improves the speed tracking ability of the system and improves the efficiency of wind energy capture.

关 键 词：风力发电系统 非线性 最大功率捕获 径向基函数神经网络 反推控制 

分 类 号：TM351[电气工程—电机]