Bearing Element Fault Diagnosis Using Support Vector Machine

Rolling bearings are the core of rotating machinery. The status of these bearings will determine the operational performance of the machine. The extraction and analysis of the fault data involved in the vibration signal of the rolling bearing are common and effective method in mechanical fault diagnosis. Therefore, identifying and judging the faults locations of rolling bearings in a timely manner can effectively guarantee the process safety of the mechanical structure, which is of great significance. In this paper, support vector machine technique is proposed without the need of feature extraction in fault diagnosis for the rolling bearing element. By using date acquired from Case Western Reverse University, SVM shows superior to convolutional neural network in the case of less data samples. The results show that this method for bearing fault identification has higher accurate rate which provides a theoretical support that ensures the safe operation of the bearing and fast fault diagnosis.