基于LS-SVM的非线性系统直接逆模型控制  

Straight Inverse Model Control for Nonlinear System Based on LS-SVM

作　　者：胡良谋[1] 曹克强[1] 李小刚[1] 徐浩军[1] 董新民[1] 

机构地区：[1]空军工程大学,西安710038

出　　处：《中国机械工程》2010年第13期1553-1556,共4页China Mechanical Engineering

基　　金：中国博士后科学基金资助项目(20090460116)

摘　　要：针对非线性系统逆模型建立难的问题,提出了基于最小二乘支持向量机(LS-SVM)的非线性系统逆模型辨识建模的方法,在此基础上,提出了基于LS-SVM的非线性系统直接逆模型控制的方法。仿真试验结果表明,采用LS-SVM建立的非线性系统逆模型具有很高的建模精度和较强的泛化能力,基于LS-SVM的直接逆模型控制方法的控制效果好,简单易行,从而验证了这两种方法的有效性和先进性。Aiming at the problem of hard system identification modeling for nonlinear system,a method of inverse model identification for nonlinear system based on LS-SVM was studied in detail.Then the method of straight inverse model control for nonlinear system based on LS-SVM was put forward.The simulation results show that inverse model of LS-SVM has high modeling precision and strong generalization.And the control effectiveness of straight inverse model control based on LS-SVM is good and is very easy to be actualized.Therefore the identification method and the control method mentioned above are valid and advanced.

关 键 词：最小二乘支持向量机(LS-SVM) 非线性系统 逆模型辨识 直接逆模型控制 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]