Active Noise Cancellation of Rail Vehicles Based on a Convolutional Fuzzy Neural Network Prediction Approach

Focusing on the problem of nonlinear noise cancellation caused by complex rail vehicles noise sources, this study proposes an active noise cancellation system based on the noise prediction of a convolutional fuzzy neural network. The multi-source noise signals of rail vehicles are predicted by the convolutional network to obtain the virtual error signal of the ideal noise reduction point, and then a fuzzy neural network is used to reduce the noise in a targeted manner. The simulation indicated that the system can operate stably and can effectively control the low-frequency noise of multi-source rail vehicles below 1000 Hz.