Zhong Yang
Institute of Mechanical and Electrical Engineering, Jinling Institute of Technology, Nanjing, 211169, China
Haifei Si
Institute of Mechanical and Electrical Engineering, Jinling Institute of Technology, Nanjing, 211169, China
Hui Zhao
Institute of Mechanical and Electrical Engineering, Jinling Institute of Technology, Nanjing, 211169, China
ABSTRACT
A rule-based expert system is a traditional approach in condition monitoring and diagnostics for complex system. However, the rule-based expert system is not only difficult to be established but also difficult to be renewed along with the changed circumstances. Neural networks provide a data based approach to condition monitoring and diagnostics for complex system such as rotating machinery. By developing associations between neural networks and a rotating machine consisting of gears, bearings and shafts for the first time, a number of advantageous aspects are identified in this study. Fundamental and harmonic frequencies relating to the components, as well as sideband and cepstrum information, were used as input parameters. Outputs of the networks were given as severity levels of system components. Neural networks demonstrated the capability for use in identifying the location and severity of numerous different machinery faults, including multiple component faults. And neural network is not just easy to be established but also easy to be renewed along with the changed circumstances.
PDF References Citation
How to cite this article
Zhong Yang, Haifei Si and Hui Zhao, 2013. Condition Monitoring and Diagnostics for Complex System using Neural Networks. Journal of Applied Sciences, 13: 2710-2714.
DOI: 10.3923/jas.2013.2710.2714
URL: https://scialert.net/abstract/?doi=jas.2013.2710.2714
DOI: 10.3923/jas.2013.2710.2714
URL: https://scialert.net/abstract/?doi=jas.2013.2710.2714
REFERENCES
- Liu, T.I. and J.M. Mengel, 1992. Intelligent monitoring of ball bearing conditions. Mech. Syst. Signal Process., 6: 419-431.
CrossRefDirect Link - Yang, Z. and H.F. Si, 2012. Integration of neural network and symbolic inference and its application to detection of foundation pile. Adv. Mater. Res., 383-390: 4977-4981.
CrossRefDirect Link - Hui, Z., C.S. Poon and T.C. Ling, 2013. Properties of mortar prepared with recycled cathode ray tube funnel glass sand at different mineral admixture. Constr. Build. Mater., 40: 951-960.
CrossRef