Application of RBF Neural Network Optimized Based on K-Means Cluster Algorithm in Fault Diagnosis

摘要

Radial Basis Function(RBF) neural network based on K-means cluster algorithm is widely used in intelligent fault diagnose with its good performance for nonlinear problems. However, the selection of initial center and number of hidden layer neurons is random. In this paper, a neural network based on improved K-means cluster algorithm with data density is proposed to solve this problem. The improved algorithm is applied to synchronous condenser's historical data. Simulation results prove the feasibility of the improved algorithm.