Motor Bearing Fault Diagnosis Method Based on Gramian Angle Field Image Coding and CNN-SVM

The most common failure in motors is bearing failure. Traditional fault diagnosis methods require manual feature extraction, which has a high level of uncertainty and complexity. This article proposes a motor bearing fault diagnostic method based on Gramian angular field image encoding and CNN-SVM. The method consists of three stages: Firstly, the original vibration signals of the motor bearing are transformed into two-dimensional images using Gramian angular field image encoding. Next, an enhanced CNN model is employed to accurately and quickly identify the elements in the images. Finally, a Support Vector Machine (SVM) is used as the final classifier to further improve the accuracy and speed of fault classification. In the experimental design, the proposed model is validated and analyzed using a bearing fault dataset from Case Western Reserve University. The results demonstrate the feasibility and superiority of this method, providing a theoretical basis for guiding bearing maintenance decisions.