带自适应模糊神经网络补偿的自耦合PI在风力机变桨控制中的应用  

Application of self-coupled PI with adaptive fuzzy neural network compensation in wind tur-bine pitch control

作　　者：何胜华 黄弘 范必双[1] 怀晓伟 He Shenghua;Huang Hong;Fan Bishuang;Huai Xiaowei(Changsha University of Science and Technology,Changsha 410004,China;SPIC Jiangxi Electric Power Co.,Ltd.,Nanchang 330001,China;State Key Laboratory of Disaster Prevention and Reduction for Power Grid Transmission and Ditrubution Equipment,Human Disaster Prevention Technology Co.,Ltd.,Changsha 410131,China)

机构地区：[1]长沙理工大学,湖南长沙410004 [2]国家电投集团江西电力有限公司,江西南昌330001 [3]湖南防灾科技有限公司(电网输变电设备防灾减灾国家重点实验室),湖南长沙410131

出　　处：《可再生能源》2023年第5期618-624,共7页Renewable Energy Resources

基　　金：国家自然科学基金项目(51877011)。

摘　　要：风力机对于风能的捕获主要依靠主控制器和执行机构之间的协同工作。然而风速的幅值和方向是时刻变化的,它的这种非线性特性给能量转换带来了挑战。针对该问题,文章设计了一种基于混合控制方案的变桨距系统,用于超风速下功率吸收的恒定。所提方法利用一种新颖的自耦合PI获取转子速度跟踪性能。而风速的随机性变化被视为一个外在扰动,采用自适应模糊神经网络对其进行补偿控制。为了验证该变桨距控制系统的性能,在Matlab/Simulink中搭建了2 MW直驱式风机模型,并与传统的PI控制器进行仿真对比。仿真结果表明,文章所提算法在转子速度跟踪和输出功率恒定控制等方面拥有比传统PI更为优异的性能。The capture of wind energy by wind turbines relies on the collaboration between the main controller and the actuators.However,wind speed is a physical variable that varies in magnitude and direction from time to time,and its nonlinear nature poses a challenge for energy conversion.To address this problem,this paper designs a hybrid control scheme-based pitch system for constant output power at super wind speeds.The proposed method utilizes a novel self-coupled PI to obtain the desired rotor speed tracking performance.In contrast,the random variation of wind speed is considered as an extraneous disturbance and an adaptive fuzzy-neural network is used to compensate for its control.To verify the performance of this pitch control system,a comparative simulation with the conventional PI algorithm is guided on a 2 MW direct-drive wind turbine model built in Matlab/Simulink.The simulation results show that the proposed scheme possesses a better response performance than PI in terms of rotor speed and output power.

关 键 词：直驱式风机 变桨距控制系统 自耦合PI 自适应模糊神经网络 补偿控制 

分 类 号：TK81[动力工程及工程热物理—流体机械及工程]