LS-SVM构造FLANN的非线性自适应逆控制动态补偿研究  

Research on Nonlinear Adaptive Inverse Control Using FLANN Data-together Constructed by LS-SVM

作　　者：孟萍[1] 张金敏[1] 张绘敏[1] 

机构地区：[1]兰州交通大学自动化与电气工程学院,甘肃兰州730070

出　　处：《兰州交通大学学报》2010年第6期70-73,共4页Journal of Lanzhou Jiaotong University

摘　　要：在对常规函数链接型神经网络(FLANN)构造方法认识的基础上,研究了一种基于最小二乘支持向量机(LS-SVM)构造FLANN的新方法,并利用该方法对非线性对象模型及逆模型进行建立.将该方法的非线性系统辨识技术应用于自适应逆控制中,提高非线性系统的自适应性,改善动态特性.设计出了一种自适应逆控制系统,不仅可以得到较好的动态响应,还能使扰动减小到最小.With the recognition of the construction method of generic functional link artificial neural network（FLANN）,a novel construction method based on least squares support vector machine（LS-SVM）is discussed and applied to the method of nonlinear object model and inverse model building.The method of nonlinear system identification technology is applied to the adaptive inverse control to increase the self-adaptive nonlinear system and improve the dynamic properties.The self-adaptive inverse control system can not only get a better dynamic response but also reduce the noise and disturbance to a minimum.

关 键 词：函数链接型神经网络 最小二乘支持向量机 自适应逆控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]