Improved Particle Swarm Optimization Algorithm with Fireworks Search

Particle Swarm Optimization (PSO) is an iterative swarm intelligence algorithm, whose main idea is introducing random factors into the evolutionary equation to simulate the foraging behavior of birds. However, this random search model gives rise to the weakened local search ability of PSO, the premature convergence problem and the inability to find the optimal solution. To improve this issue, an Improved Particle Swarm Optimization with Fireworks Search (FS-PSO) is proposed in this paper. In FS-PSO, explosive sparks and gaussian mutation sparks are used to search at the end of the iteration of PSO based on the search mechanism of the Fireworks Algorithm (FWA), which is aiming to increase the population diversity of the particles and enhance the local search ability of PSO. The simulation results show that the improvement of FS- PSO is effective to the premature convergence problem of PSO.