Affiliation:
1. Zhejiang Technology Institute of Economy
2. Zhejiang Institute of Mechanical & Electrical Engineering
3. Zhejiang University
Abstract
his paper reports a practical approach for detecting and diagnose engine faults in real-time based on both the historical and the real-time engine operation data using a specially design neural networks-based fault diagnosis expert system. This system consisted of multiple sensors for real-time monitoring, an engine database for historic data comparison, and a neural network-bases classifier for detecting faults based on both the real-time and the historic data. This neural network-based engine fault diagnosis system was evaluated in a series of validation tests. The results indicated that the system was capable to detect the predefined faults reliably, and the diagnosis error was less than 5%.
Publisher
Trans Tech Publications, Ltd.
Subject
Mechanical Engineering,Mechanics of Materials,General Materials Science
Reference10 articles.
1. Wollenberg, B.: IEEE T Power Syst. vol. PWRS-1(1986), p.241.
2. Fauquembergue, P., Perrot, L.: 2nd Int. Symp. Expert Syst. Applicat. Power Syst., Seattle, WA, 1989, p.264.
3. Fukui, C., Kawakami, J.: IEEE T Power Deliver. vol. PWRD-1(4). (1986), p.83.
4. Germond, A., J., Niebur, D.: Europe. Proc. IEEE, vol. 80(1992), p.732.
5. Komai, K.: IASTED Int. Symp. High Technol. Power Ind., Bozeman, MT. (1986), p.366.
Cited by
3 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献