A Quadratic Particle Swarm Optimization and its Self-Adaptive Parameters

Particle swarm optimization (PSO) is a kind of random optimization algorithm based on the swarm intelligence. This paper presents a quadratic PSO by improving the standard PSO´s evolution equation on the foundation of analyzing standard PSO´s model and its mechanisms, and introduces a parameter automation strategy for it on the basis of comparing quadratic PSO with PSO and analyzing the impact that the parameters have on the performance of algorithm. The simulation illustrates that the new method improved the performance of the PSO. Further, for most of the benchmarks function, the self-adapting parameters strategy outperformed the fixed parameters. The experimental results show that the quadratic PSO is feasible and the strategy is correct and efficient