Research on Fault Diagnosis for Rotating Machinery Vibration of Aero-Engine Based on Wavelet Transformation and Probabilistic Neural Network
Fault feature extraction using wavelet decomposition and probabilistic neural network fault diagnosis technology is presented in this paper. Fault diagnosis based on wavelet transformation and neural network data fusion is studied. The fault diagnosis in rotating machinery vibration of the aero-engine is simulated in Matlab. Our recent investigations demonstrate that using wavelet decomposition extract fault characteristics of the energy vector has strong generalization ability and anti-noise ability. Integration of the wavelet and neural network application can provide a better classification of diagnosis results, reliability and accuracy. This technique is suitable for the mechanical vibration fault diagnosis applications of steam turbine and gas turbine.
Pengcheng Wang, Liqun Ai, Yungang Li, Xiaoming Sang and Jinglong Bu
W. J. Wu and D. G. Huang, "Research on Fault Diagnosis for Rotating Machinery Vibration of Aero-Engine Based on Wavelet Transformation and Probabilistic Neural Network", Advanced Materials Research, Vols. 295-297, pp. 2272-2278, 2011