A Comparative Study on Fault Diagnosis of Electrical Vehicle Motor Testing Machine Gear Components Using Machine Learning Algorithms

<div class="section abstract"><div class="htmlview paragraph">Electric vehicles (EVs) are paving the way for future mobility, with drive motors playing a central role in their efficiency and performance. Motor testing machines are crucial for validating EV motors, yet flaws in testing equipment, such as gear issues, often lead to operational disruptions. This study aims to enhance motor testing by implementing machine learning and vibration signal analysis to detect gear faults early. Using statistical feature extraction and classifiers like Quadratic SVM and Bagged Trees, the collected vibration signals are categorized as normal or faulty under loaded (0.275 kW) and no-load conditions. Performance comparison reveals the Bagged Trees algorithm's superior accuracy of 95.3%. This approach offers an intelligent, preventive maintenance solution, improving the motor test bench’s reliability.</div></div>