Fault Diagnosis of Tunnel Fan Motor Bearings Based on Three-dimensional Vibration Joint Prediction and Auto-Encoder Mechanism

Fault diagnosis of tunnel fan motor bearings is crucial to ensure the safe operation of transportation infrastructure. Existing bearing fault diagnosis methods have problems such as limited data dimension and insufficient model learning ability. To address this problem, this paper proposes a diagnostic model that integrates three-dimensional vibration joint prediction and Auto-Encoder mechanism. The model is tested on the open source dataset released by MUET. The experimental results show that the model performs significant advantages under complex working conditions and is superior to other models. Specifically, compared with the multilayer perceptron classifier, the precision, recall and F1-score are improved by 27.94%, 28.19% and 28.06% in absolute percentage, respectively.