Steam turbine vibration fault diagnosis based on particle swarm optimization clustering

A novel fuzzy clustering algorithm: PFCM was proposed based on the fusion of particle swarm optimization (PSO) and fuzzy c-means clustering (FCM). The conventional FCM has the two drawbacks of sensitivity to initialization and easily being trapped into local optima, due to the gradient descent approach used. With the features of global optimization and fast convergence, the hybrid algorithm presented can overcome these shortcomings and yield the optimal clustering performance. The new data clustering technique provided was also applied in the vibration fault diagnosis of steam turbine. Computer simulations demonstrate that compared with FCM, the proposed PFCM has a superior fault diagnosis capability.