An Improved Particle Swarm Optimization Algorithm with Immunity

Abstract

The particle swarm optimization (PSO) algorithm is a relatively new kind of intelligent optimization algorithm. PSO is a stochastic, population-based optimization technique that is based on a metaphor of social behavior, namely bird flocking or fish schooling. Although the algorithm has shown some important advances, such as easier implementation, fewer presetting parameters and higher speed of convergence, it has also been reported that the algorithm has a tendency to get stuck in local optimum and may find it difficult to improve solution accuracy by fine tuning. This is due to a decrease of diversity during the evolutional process that leads to plunging into local optimum and ultimately fitness stagnation of the swarm. In order to maintain appropriate diversity and rapid convergence, an improved PSO algorithm with immunity is proposed in the paper. Immune memory and immune vaccination are adopted in the proposed PSO algorithm (shorten as IVPSO). The diversity of population is extended adequately, and the risk of premature convergence is depressed effectively in IVPSO algorithm. Testing over the benchmark problems, the experimental results indicate the IVPSO algorithm prevents premature convergence to a high degree and has better convergence performance than standard PSO algorithm.