Abstract: With the rapid development of intelligent system, real time optimization become more and more urgent. Particle Swarm Optimization (PSO) is one of the most effective algorithms in solving such problems. Considered the complexity of intelligent system optimization, speed-up technique is needed. As many optimization problems can be converted to travelling salesman problem, the standard benchmark problem of TSP with 31 cities is employed to analyze the relationship between optimal solution and different parameters. The effect on average of the optimal solution, optimal solution, convergence speed and stability of the optimal solution of different parameters are analyzed. Finally, a comparation with ant colony algorithm is conducted and suitable values of parameters are proposed.