PID Parameters Turn and Simulate Based on RBF Neural Networks

摘要：

This paper adopts RBF Neural Networks to control the nonlinear and time-varying object accurately. The function of RBF is regarded as the base in middle layer. The input vectors are mapped to dormant space. The adjustable parameters in the networks are the weight of linear combination because the mapping is linear from connotative layer to the space of output solution. This method is adopted to simulate to control the single-input and single-output object. The simulative result shows that this method has the adaptive characters.

关键词： Radial Basis Function Networks PID Control Neural Networks Nonlinear Time-varying object

作者: Hui Li Dejiang Zhang Jun Lin Wenxue Li

作者单位: School of Electrical & Electronic Engineering, Changchun University of Technology, Changchun, Jilin Key Lab of Geo-lnformation Exploration & instrumentation, Jilin University, Changchun,Jilin 130026, School of Electrical & Electronic Engineering, Changchun University of Technology, Changchun, Jilin

会议类型: 国际会议

会议名称: 2006 International Symposium on Distributed Computing and Applications to Business,Engineering and Science(2006年国际电子、工程及科学领域的分布式计算应用学术研讨会)

会议地点: 杭州

会议语种:英文

页码: 1020-1022

在线出版日期: 2006-10-12（万方平台首次上网日期，不代表论文的发表时间）