A new ensemble fault diagnosis method based on K-means algorithm

Abstract

A new ensemble algorithm based on K-means clustering and probabilistic neural network called K-means-PNN for classifying the industrial system faults is presented. The proposed technique consists of a preprocessing unit based on K-means clustering and probabilistic neural network. Given a set of data points, firstly the K-means algorithm is used to obtain K-temporary clusters, and then PNN is used to diagnose faults. To validate the performance and effectiveness of the proposed scheme, K-means-PNN is applied to diagnose the faults in TE Process and compared with K-means clustering and back-propagation neural network called K-means-BP, algorithm. Simulation studies show that the proposed K-means-PNN algorithm compared with K-means-BP algorithm not only improves the accuracy in fault classification but also is a reliable and computationally efficient tool.