Abstract：A multi-scale cooperative searching simplified PSO( MSPSO) was put forward to avoid the standard PSO( SPSO) premature,which was put forward by simulating ants working cooperatively in nature. Its particles trajectories were analyzed and its convergence was proved. To test and compare MSPSO and SPSO algorithm performance,VB6. 0 was used to do numerical experiment with three commonly used test function. The numerical experiment indicated that the MSPSO has better searching performance,especially in solving the higher dimensional optimization problems.

Keyword：particle swarm optimization;multi-scale;convergence;global search;