Fault diagnosis of rotating machinery based on MFES and D-S evidence theory

In real applications of rotary machinery, sometimes multiple-faults may occur and the fault diagnosis based on single sensor with limited information may be low reliability. Therefore, an approach of multiple-faults diagnosis for rotor-bearing systems based on multiple frequency energy spectrum (MFES) and Dempster-Shafter (D-S) evidence theory is presented in this paper. Firstly, the original acceleration signals are processed by fast Fourier transformation (FFT) from the time domain to frequency domain. According to the analysis of the frequency information, the MFES is put forward to extract features from vibration under normal and faulty conditions of rotational mechanical systems. In order to get the best MFES, the impact factor η decide energy interval in frequency is studied and five kinds of features with different η are calculated. Secondly, these features were given as inputs for training and testing the model of the RBF neural network. Finally, the all RBF neural networks results of multi-sensors are fused by D-S evidence theory, and the result is counted the final diagnosis conclusion. Experimental results show that the method is effective and feasible for fault diagnosis of multiple-faults.