Battery life time prediction of electric vehicle using artificial intelligence

Abstract

The widespread use of electric vehicles (EVs) is viewed as a turning point for lower emissions of co2 and much more advanced driver assistance systems. It is time-consuming to predict the battery capacity of the electric vehicle using the endurance and dependability of battery systems. Because battery deterioration is typically non-linear, predicting the capacity of the battery of charge estimation with significantly less deterioration is incredibly time-consuming. As a result of this complexity, real-time control of battery storage has proven difficult. Nevertheless, using the latest advancements in battery deterioration comprehension, simulation techniques, and testing, there is a potential to combine this expertise with new machine learning (ML) approaches to find a potential solution for this complexity. In this study, the battery life is predicted using three different regression models Linear Regression, Ridge Regression, and Ensemble Regression. The model's error analysis is conducted to improve the battery's performance characteristics. Finally, the three ML algorithms are compared using performance parameters. Out of the three regression models, the Ensemble regression model is found to be the best model which has an accuracy of 94.8%.