Adaptive rotating machinery fault diagnosis method using MKIST

Abstract

Abstract An adaptive fault diagnosis method for rotating machinery based on maximum kurtosis incomplete S-transform is proposed in this paper. Firstly, the incomplete S-transform is performed on the fault frequency band of the vibration signal, and the module vector group is obtained through module calculation. Subsequently, the kurtosis of all the modulus vectors are calculated and the vector corresponding to the maximum kurtosis is located to adaptively determine the envelope of the fault frequency component in the vibration signal. Then, fast Fourier transform is performed on the envelope to obtain its main frequency, which is matched with the fault mode frequency to achieve fault diagnosis of rotating machinery. Finally, the mean peak ratio (MPR) was used to evaluate the performance of different methods under various operating conditions. The results show that the maximum MPR is obtained by the proposed method, demonstrating its stronger noise resistance and demodulation ability.