Time delays identification by means of a hybrid interpolated ant colony optimization and Nelder-Mead algorithm

Abstract

The paper presents a possibility of application a hybrid algorithm compound of a global algorithm Interpolated Ant Colony Optimization and local Directional Derivatives Simplex (IACO-DDS) to identification of the time delays in the linear plant. The Interpolated Ant Colony Optimization algorithm (IACO) can be used to continuous problems. The Derivatives Simplex belongs to the class of a modified Nelder Mead Algorithm. The IACO-DDS algorithm is compared with the simple IACO in terms of their speed and efficiency for unimodal and multimodal search space. The investigation was made for the sixteen plants with random parameters, divide into two group: the simple and difficult to identification plants.