基于神经网络的静止无功补偿器自校正内模控制  

NEURAL NETWORK BASED SELF-TUNING INTERNAL MODEL CONTROL FOR SVC

作　　者：彭建春[1] 王耀南[1] 黄纯[1] 

机构地区：[1]湖南大学电气系,长沙410082

出　　处：《电网技术》1997年第11期32-36,共5页Power System Technology

摘　　要：本文利用BP神经网络逼近受控系统的动态及其逆动态，设计了一种静止无功补偿装置（SVC）的自校正内模控制器。该控制器的正模型和逆模型都以三层BP神经网络为主体，实现对SVC及电网的动态描述和对SVC的控制。所设计的控制器不需要电力网络及SVC的数学模型，并且具有良好的鲁棒性和控制精度。A neural network based self-tuning internal model controller for SVC is presented in this paper. Three-layer BP neural network models are used for the internal model and the reverse model of the controlled system in the controller. The controller needs no models of SVC and power network and is of good robustness as well as control accuracy.

关 键 词：静止无功补偿器 神经网络 自校正控制 电网 

分 类 号：TM714.3[电气工程—电力系统及自动化]