Comparison of the convergence of IIR evolutionary digital filters and other adaptive digital filters on a multiple-peak surface

Abstract

This paper demonstrates a comparison of the convergence behavior of the IIR evolutionary digital filter (IIR-EDF), the LMS adaptive digital filter (LMS-ADF) and the adaptive digital filter based on the simple genetic algorithm (SGA-ADF) on a multiple-peak surface. In numerical examples, the authors use a reduced-order system identification to simulate a multiple-peak surface in which local minimum problems can be encountered. The experimental results show that the EDF adaptive algorithm can search the global minimum in the multiple-peak surface of these examples and has a smaller adaptation noise than the other algorithms.