Abstract:  To solve the problem that weak signal and strong signal of planetary gearbox fault were difficult to be separated at low speed, which led to poor diagnosis accuracy of planetary gearbox weak fault, a planetary gearbox fault diagnosis method based on local mean decomposition (LMD) was proposed. Firstly, data acquisition & signal processing(DASP)data acquisition system was used to collect vibration signals of planetary gearbox under different working conditions, and the vibration signals were denoised by translation invariant wavelet denoising method. Then the vibration signal was decomposed according to the local mean decomposition method, and the energy operator and cycle frequency were respectively used to demodulate the vibration signal. The amplitude and phase modulation information corresponding to the weak fault signal components were obtained, and the characteristics of the weak fault signal of the planetary gearbox were accurately extracted. Finally, least square support vector machine (LSSVM)was used to identify different fault features, judge the running state of planetary gearbox, and realize the fault diagnosis of planetary gearbox. The research results show that the proposed method can accurately diagnose the weak abnormal signals and strong abnormal signals of planetary gearboxes, and obtain satisfactory fault diagnosis results of planetary gearboxes. It can effectively ensure the safe and stable operation of planetary gearboxes when applied in practical production.
Key words:  gear transmission; local mean decomposition(LMD); least square support vector machine(LSSVM); translation invariant wavelet noise reduction; vibration signal noise reduction; weak fault signal characteristics