An Improved Particle Swarm Optimization Algorithm for FIR Filter Design

In this paper, an Improved Particle Swarm Optimization algorithm is introduced. Compared with conventional PSO, a new inertia weight mechanism is used to ensure the coverage and convergence at a better extent, considering the suboptimal solutions generated by conventional PSO when dealing with complex problems with lots of local minimas. The new mechanism divide the particles into two parts to enlarge coverage of search space and apply new decreasing algorithm to guarantee convergence. FIR filter design, whose results are very sensitive to parameters, is used to illustrate the effective impact of the new inertia weight mechanism.