In order to enhance the global convergence ability and the convergence speed of particle swarm optimization (PSO), based on the results of reference, an improved strategy is proposed in this paper. Inspired by the effect of individual experience in animal groups on the move, in the method, the whole searching process is departed into two steps: the exploration search with leadership and the exploitation search. The mutation operator is used to add the diversity of swarm and avoid the convergence of particles to local optimum solutions prematurely. Through analyzed five benchmark functions, the improved method is compared with two famous PSO methods. And, the variance analysis of statistic theory is applied to compare the performance of the three methods. Experimental simulation results confirm that the proposed algorithm can not only significantly speed up the convergence, but also effectively solve the premature convergence problem.