Abstract: Because fuzzy clustering algorithm is sensitive to initial cluster center and easy to fall into local optimum,improved fuzzy clustering algorithm based on parallel niches particle swarm optimization algorithm is presented in this paper.Valley function is used to form mutually exclusive subswarms by identifying niches.Punishment function is adopted to realize information sharing mechanism in the process of subswarms parallel searching.Hybrid cluster validity function is imported to obtain optimal cluster number.Simulation results show that the improved algorithm can enhance searching efficiency and classification accuracy of fuzzy clustering algorithm.