Multimodal Deep Learning For Multiple Motor And Sensor Faults Diagnosis

抄録

Condition monitoring and fault diagnosis is of significance to improve the safety and reliability of motor system. This paper presents a novel approach for motor fault diagnosis based on multimodal deep learning. Particularly, the case that the motor fault is coupled with the sensor fault has been taken into consideration. One-dimensional convolutional neural network and multi-layer perceptron are employed in the proposed model to extract features from vibration signals and current signals respectively, and a fully connected layer gets the shared representation of various modal. The proposed model has been applied to analyze experimental signals collected from different motor conditions. Results show that the proposed model is effective in diagnosing motor faults and sensor faults for improved induction motor status assessing and maintenance.