A Study on Fault Classification System Based on Deep Learning Algorithm Considering Speed and Load Condition

Many types of research are underway on unmanned fault diagnosis of plant facilities. Deep learning algorithms are generally used for this purpose. Previous studies have been carried out assuming multiple motor fault conditions considering constant speed. However, the motor speed can be changed to the user purpose. In this paper, a study to develop a deep learning fault classification algorithm that considers the motor driving conditions was carried out. Especially a data preprocessing algorithm was used to shift the data in the frequency domain to account for the driving conditions. In addition, the proposed model became capable of classifying various faults. As a result, the fault classification algorithm considering the motor driving conditions showed enough performance of 98% or more. In addition, aging motor and similar fault diagnosis were also discussed.