Optimal time delay estimation for system identification

Abstract

This paper presents a novel strategy for optimal time delay estimation for system identification purposes. A brief review of classical time delay estimation techniques is presented and the problem of time delay estimation is further inspected from a system identification perspective. A solution based on optimization of the fit index is then proposed to find the best time delay, so that a given model structure with a certain order and adjusted for an identification dataset best represents the true system. A set of examples illustrates and validates the application of the proposed technique. © 2013 AACC American Automatic Control Council.