A Hybrid Particle Swarm Optimization Algorithm for Multimodal Function Optimization

Particle swarm optimization (PSO) has shown its good performance on numerical function problems. However, on some multimodal functions the PSO easily suffers from premature convergence because of the rapid decline in velocity. In this paper, a hybrid PSO algorithm, called HPSO, is proposed, which employs a modified velocity model to guarantee a non-zero velocity. In addition, a Cauchy mutation operator conducted on the global best particle is used for improving the global search ability of PSO. Experimental studies on a suite of multimodal functions with many local minima show that the HPSO outperforms the standard PSO, CEP, Gaussian swarm with Gaussian mutation (GPSO+GJ) and Gaussian swarm with Cauchy mutation (GPSO+CJ) on most test functions.