Fault diagnosis and prognosis of machinery has become more important than ever as modern industry develop. In this paper we present a novel fault prognosis approach using echo state network (ESN) and recurrent multilayer perceptron (RMLP), two functional forms of recurrent neural network, based on vibration signal of rotating machinery. Both ESN and RMLP methods are able to improve the capacity of machinery performance prediction within a relatively short time period and only with limited data available. The successful outcomes of the ESN/RMLP fault prognosis approach can give obvious explanation for future states of machine, which make it possible to enhance machine condition monitoring and health management in practical situations. The experiment results show that accuracy of fault prognosis has improved significantly, and prove that the proposed approach is practical and efficient for fault prognosis of the rolling bearing system.
The application of echo state network and recurrent multilayer perceptron in rotating machinery fault prognosis
2016-08-01
302790 byte
Conference paper
Electronic Resource
English
Modular, Multilayer Perceptron
NTRS | 1991
|Convolutional Neural Network Based Fault Detection for Rotating Machinery
Online Contents | 2016
|Fault Diagnosis of Rotating Machinery Using Back Propagation Neural Network
British Library Online Contents | 1997
|