基于小波网络的非线性系统建模与控制  

Modeling and Control of Nonlinear Systems\= Based on Wavelet Networks

作　　者：王群仙[1] 陈增强[1] 袁著祉 

机构地区：[1]南开大学计算机与系统科学系

出　　处：《控制与决策》1999年第4期359-363,共5页Control and Decision

基　　金：国家863计划应用基础研究基金;天津市自然科学基金

摘　　要：提出一种基于小波网络的非线性系统的建模和控制方法。使用小波网络对未知控制系统建立一步预测模型，基于Ｄａｖｉｄｏｎ最小二乘法得到自适应控制律。小波网络的权值由广义递推最小二乘法来学习，尺度参数和平移参数通过稳定的Ｄａｖｉｄｏｎ最小二乘法来获得。仿真结果表明了该方法的有效性。This paper presents a method of modeling and control for nonlinear systems based on wavelet networks. This method performs single-step modeling using wavelet network for unknown controlled systems and Davidon least squares algorithm optimizes the control law. The weights in wavelet network are learned by general recursive least squere method and dilation with translation parameters by Davidon least square algorithm. Simulation results shows that the controller can be successfully applied in nonlinear systems.

关 键 词：小波网络 预测控制 非线性系统 建模 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]