System identification by Crazy-cat swarm optimization

摘要

Adaptive filtering and system identification by traditional derivative based algorithms create stability issues when used in infinite impulse response (IIR) systems. In this paper, the identification of IIR system is used as an optimization task. A modification is approached to cat swarm optimization by introducing the concept of craziness to produce Crazy cat swarm optimization(Crazy-CSO) algorithm. The new modified version of the algorithm has been utilized to find a better solution. The efficiency of the modified algorithm is verified by identification of few standard IIR systems through simulation study. The new method exhibits finer identification performance as compared to particle swarm optimization (PSO) and cat swarm optimization (CSO) based identification by providing superior outputs.