Using PSO method for system identification

Abstract

The article deals with application of heuristic method in system identification process. Real measured data can be used for system identification. Usually real data has noisy component. For this reason sometimes the filtering is applied to original data and the result is used for identification. In this article the comparison of usage the original and prefiltered data is discussed. For identification the Particle swarm optimization method is used. Different approaches are also used to create initial agents' positions. Whole concept is tested using the MATLAB program. © Springer International Publishing Switzerland 2014.