基于自适应神经网络的高阶非仿射变速风力机系统全局跟踪控制方法  

UNIVERSAL TRACKING CONTROL METHOD FOR HIGH-ORDER NON-AFFINE VARIABLE SPEED WIND TURBINE SYSTEM BASED ON ADAPTIVE NEURAL NETWORK

作　　者：王磊[1] 陈晨[1] 沈涛[1] 

机构地区：[1]重庆大学自动化学院,重庆400030

出　　处：《太阳能学报》2017年第8期2083-2089,共7页Acta Energiae Solaris Sinica

基　　金：国家自然科学基金(51205046);重庆市研究生科研创新项目(CYS14007)

摘　　要：根据高阶非仿射变速风力机的特性,提出自适应神经网络全局跟踪控制方法,实现对变速风力机转速的跟踪控制,解决由风的波动性和随机性导致的变速风力机控制中长期存在的非线性时变参数问题。首先,建立变速风力机的非线性模型,通过系统转换,将非仿射系统模型转换为仿射的严格反馈系统模型。然后,使用神经网络观测器对系统不可得到的状态进行估计。最后,将自适应神经网络全局跟踪控制方法用于系统的局部控制器,并通过仿真验证该控制方法具有良好的控制效果。According to the characteristics of high-order non-affine variable-speed wind turbine （VSWT）, the universal tracking control method based on adaptive neural network was proposed to track and control the speed of VSWT, and solve long-term nonlinear time-varying parameter problem existing in the control of VSWT caused by wind volatility and randomness. At first, the nonlinear model of VSWT was set up, the model of non-affine system was transferred into the model of strict affine feedback system through system transfer. Then neural network observer was used to estimate not available state of the system. Finally, universal tracking control method based on adaptive neural network was used for local controller of the system. The simulation is carried out and the results verify that the control method has good control effect.

关 键 词：自适应神经网络 跟踪控制 变速风力机 风能 

分 类 号：TM721[电气工程—电力系统及自动化]