Among all kinds of gear fault diagnosis methods, the gear fault diagnosis method based on vibration detection has the advantages of simple measurement and strong real-time. Through measuring the vibration signal generated in the process of gear operation, as an important information source of fault diagnosis, it is an ideal online operation monitoring tool for gear transmission state. The key of vibration detection and fault diagnosis is how to extract and separate the weak information related to gear fault features from the complex vibration signals.
Spectrum analysis is a description of the distribution of the original signal in the frequency domain, which can usually provide more intuitive feature information than the waveform in the time domain. Therefore, spectrum (including power spectrum and amplitude spectrum) is widely used as the basis of fault diagnosis. The spectrum can be obtained by Fourier transform. It is worth mentioning that in the vibration spectrum of the machine, although some vibration components are large, they do not change with time and pose no threat to the normal operation of the machine. On the contrary, there are some small amplitude, but the fast-growing frequency component often indicates the generation and development of fault, which should be paid enough attention.