Parameter selection for particle swarm optimization based on Stochastic multi-objective optimization

Parameter selection in particle swarm optimization algorithm had great influence on its performance. This study presented a method of parameter optimization for the particle swarm optimization algorithm based on Stochastic multi-objective optimization. Based on the analysis of the relationship of inertia weight, cognitive coefficient and social coefficient, a stochastic convergence speed index function and a stochastic convergence precision index function for the standard particle swarm optimization algorithm model were analyzed and established, where some of the parameters were stochastic. Then, as two targets of these index functions in a multi-objective optimization problem, a kind of multi-objective optimization method was applied to optimize the parameters. By optimizing the two index functions, a Pareto set and corresponding parameter selection guidelines were derived to guarantee that the particle swarm optimization algorithm had a good convergent speed and accuracy in the stochastic sense. In practical application, it provided some theoretical basis, and had a helpful guiding significance.