Fault Diagnosis Method Based on Improved Particle Swarm Optimization Clustering

In order to overcome the Fuzzy C-means Algorithm’s defects of sensitivity to the initial cluster centers, using the efficient global optimization characteristics of the PSO algorithm, this paper proposes a new PSO-based fuzzy algorithm (PSO-C-FCM) . It first finds the optimal extreme using the PSO algorithm, and then initializes the cluster centers of FCM algorithm with the optimal extreme, which makes the algorithm more efficient and accurately. This new algorithm is applied to the motor fault diagnosis, and experiments show that the algorithm makes up the defects of Fuzzy C-means Algorithm, improves the efficiency and accuracy of fuzzy clustering, and improves the fault diagnosis.