Engine Gearbox Fault Diagnosis Using Modified Elman Neural Network and ACO Algorithm

• Published in: Applied Mechanics and Materials Vol. 190-191; pp. 982 - 986
• Main Authors: Song, Chong Zhi, Wu, Hai Chao
• Format: Journal Article
• Language: English
• Published: Zurich Trans Tech Publications Ltd 01.07.2012
• Subjects: Engine Gearbox; Modified Elman Neural Network; Ant Colony Optimization (ACO); Fault Diagnosis; Neural Network (NN)
• ISBN: 9783037854624; 3037854626
• ISSN: 1660-9336; 1662-7482; 1662-7482
• DOI: 10.4028/www.scientific.net/AMM.190-191.982

Analyzed the shortcomings of Elman network, the paper put forward a modified Elman network, combined Ant Colony Optimization (ACO) algorithm to train the modified Elman network, and implement trained NN (Neural Network) to fault diagnosis of engine gearbox. Using conventional ’frequency domain’ analysis method, modified Elman network fault diagnosis of the gearbox was carried out. The results proved that fault diagnosis of engine gearbox based on the modified Elman neural network and ACO has better precision and diagnoses gearbox effectively, which imp roves the effectiveness and quality of the diagnosis.