162 / 2025-06-07 15:22:29

Research On Vehicle State Recognition Based On Improved CNN

vehicle status, PHM, improved CNN, engine

Traditional fault prediction and health management (PHM) methods require complex signal processing, expert experience, and the accuracy of fault identification is low. To solve these problems, a fault diagnosis method of equipping engine based on improved CNN is proposed. Firstly, the vibration signals of equipping engine are collected and grouped. Then, the data are analyzed in frequency domain. Finally, the data are divided into training set and test set and input into improved convolutional neural network for feature extraction and model training to realize the fault identification of equipping engine. The results show that the classification accuracy reaches 98% under four working conditions of equipping engine.