Swash-Plate Bearing Fault Diagnosis Method Based on Improved One-Dimensional Convolutional Neural Network

Aiming at the need for helicopter fault diagnosis, research on fault diagnosis of the full-size helicopter swash-plate are carried on. Based on the electrical discharge etching technology, we did lots of fault seed experiments of swash-plate bearings such as outer race failure, inner race failure, and ball failure. On the basis of Fast Fourier Transform (FFT), data will be input into the one-dimensional convolutional neural network to diagnose the fault of the swash-plate bearings. The result shows that the proposed method reduces the interference of background noise, and the average fault detection rate is 92.03%, which can realize the fault diagnosis of swash-plate bearings.