A hybrid search strategy based particle swarm optimization algorithm

Particle Swarm Optimization (PSO) algorithm is widely used to deal with global optimization problems. However, it is easy to be trapped into local optimal and thus usually fall into premature convergence when encountering complicated problems, such as high-dimension and peak optimizations. To solve such problems, we propose a hybrid search strategy, derived by combining a grid searching and stochastic searching. The application of grid searching can separately search the optimal solution for each dimension, and therefore enhance searching ability. Such hybrid search strategy based Particle Swarm Optimization is called Grid-PSO algorithm. To ensure Grid-PSO performs well on global optimization problems by comparing with other optimization algorithms in literature, five benchmark functions are selected. The experimental results suggest the proposed Grid-PSO outperforms these optimization algorithms on the five benchmark functions.