Fault Diagnosis of Planetary Gearbox Based on Signal Denoising and Convolutional Neural Network

Planetary gearboxes are widely used in aerospace, marine and other important equipment for their unique advantages, and their health directly affects whether the equipment can work normally. The vibration signal generated when the fault occurs is extremely complicated, making it difficult to perform an effective diagnosis. In order to solve this problem, a planetary gearbox fault diagnosis method based on autocorrelation noise reduction combined with an improved convolutional neural network is proposed. The method firstly performs autocorrelation noise reduction on the fault signal. Secondly, the noise-reduced signal is used as the input of CNN for automatic feature extraction. The classifier is used to finally complete the intelligent diagnosis of the planetary gearbox.