Optimizing the Performance of the Wavelet-Correlation system identification method

Abstract

The Wavelet-Correlation method of system identification is derived as adaptation of the Correlation method. The Wavelet-Correlation method of identification has proven to be effective in estimating transfer functions particularly when a system is influenced by a disturbance. In order to optimize the accuracy of the transfer function estimates created by the Wavelet-Correlation method, the order of the filters to be identified, the number of offspring wavelets used to compute the wavelet functions, the scale parameters, and the type of filters were varied. Through the changing of various parameters, optimum identification can occur.