Multistage for identification of Wiener time delay systems based on hierarchical gradient approach

摘要: ABSTRACTThis paper deals with the problem of identification Wiener systems unknown integer time delay. This presents several difficulties, since it consists in estimating simultaneously delay, parameters linear dynamic subsystem and non-linear static block. To overcome these we propose a new approach to identify delay based on Hierarchical gradient principle. proposed method consists, first, decoupling original optimization into three simple separate criteria jointly optimized second, using algorithm minimize each function. The analysis gives us insight properties this develops condition that must be satisfied ensure convergence guarantee use rounding property. Simulation results are presented illustrate performance method.