An improved particle swarm optimizer for IIR digital filter design
Xuzhen Zhang, Pingui Jia, Junying Guo · 2010
This paper proposed an improved particle swarm optimization (PSO) algorithm called redistributing PSO (RPSO) for designing IIR digital filters. The proposed RPSO avoids the stagnation problem by automatically triggering particles redistributing when premature convergence is detected. Every particle is redistributed either within the whole problem space or around the mean between the global best and its current position. This mechanism helps particles escape from local convergence regions and continue progress toward true global optimum. The simulation results of low-pass and band-pass filters show that RPSO is better than PSO, quantum particle swarm optimization (QPSO), chaos particle swarm optimization (CPSO), and differential cultural (DC) algorithm with better mean performance and more stability and is an efficient method for IIR digital filter design.