IIR filter design and identification using NPSO technique

In this paper a modified version of swarm intelligence technique called Novel Particle Swarm Optimization (NPSO) technique is applied to IIR adaptive filter design problem. The proposed technique NPSO in close similarity with Real coded Genetic Algorithm (RGA), Particle Swarm Optimization (PSO) and Differential Evolution (DE) performs a structured randomized search of unknown parameters within a multidimensional search space by manipulating a swarm of particles to converge on optimal solutions. NPSO velocity equation is modified with good and bad experience components along with judiciously chosen random variables to bind the bird to fly only toward the promising zone of food. The exploration and exploitation of entire search space can be handled efficiently with NPSO along with the benefits of overcoming the premature convergence and stagnation problems. The simulation results justify the optimization superiority of the proposed NPSO over RGA, PSO and DE.