一种基于小波分解的非线性系统辨识的新方法-信息与控制.PDF

33 5 Vol. 33,No.5 2004 10 Information and Control Oct. , 2004 :2004) 石宏理, 蔡远利, 邱祖廉 ( , 710049) : NA X .( ) NA X , . ,. . : ; NA X ; ; : TP18 :A A New Approach of Nonlinear System Identification Based on Wavelet Decomposition SHI Hongli, CAI Yuanli, QIU Zulian ( School of Electronics and Inf or mation Engineering, Xian Jiaotong University , X ian 710049, China) Abstract: A new nonlinear system identification approach is proposed, which combines wavelet theory and NA X(nonlinear autoregressive with exogeneous inputs) model properly. The approach utilizes efficient approxi mation power of wavelet (multi dimensional wavelet) function to remove the complicated process of model structure determinetion. It constructs a rather general framework of identification without depending on a priori information of the system. The ecursive Least Square ( LS) algorithm can be used to estimate the parameters of the new i dentification model, which is feasible to realize online identification. The results of two simulation examples illus trate effectiveness of the new identification approach. Keywords: system identification; nonlinear autoregressive with exogeneous inputs(NA X) model; wavelet decomposition; recursive least square( LS) algorithm 1 ( Introduction , . NA MAX , . ., , , , . : Volterra .,