Fault Detection of Electric Vehicle using Machine Learning Algorithm

Abstract

This project aims to build prototype of fault detection and prediction of an electric vehicle using Naive Bayes algorithm. In this proposed technique, four machine learning algorithms are used to detect the EV fault. The algorithm which has higher accuracy is imported as a model in Raspberry Pi to detect the faults of an EV. The Trained Model detects the faults in an electric vehicle and reduces the impact of faults like fire accidents. The system consists of Raspberry Pi 3 processor and a trained Naive Bayes machine learning algorithm is imported into it as a model. This system also consists of a Vibration Sensor, Noise Sensor, Temperature Sensor to get real time values and use them as test data. The experimental results show whether the Trained Model detects the faults from the real time values and displays them in the terminal. The performance metrics of the algorithm imported into the Raspberry Pi 3 has higher accuracy than the other compared algorithms.