Prediction of Battery Failure in EVs Using Machine Learning: A Case Study

Electric Vehicles have been revolutionary in the automobile industry as they provide a more sustainable method of transportation. The battery is the heart of an EV as it stores electricity, making it environment friendly. They play a very important role in the reliability and overall functionality of an EV. Most EVs use Lithium-Ion batteries nowadays. These batteries are preferred for their long cycle lives and high energy density. Considering its importance, it is crucial to understand the reason for and predict battery failure to ensure the user’s safety and the system’s cost-effectiveness. Battery failure can be detected by monitoring the voltage, current, State of Charge and Temperature. Using a dataset containing the required parameters, a Machine Learning algorithm can be deployed to detect any failure by identifying any anomaly in the battery’s working. Various algorithms, such as K Nearest Neighbour, Support Vector Machine, and Random Forest algorithm, can be used to predict battery failure. This study aims to determine which algorithm is best suited for this model.