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.

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。

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