Battery Management System to Estimate Battery Aging using Deep Learning and Machine Learning Algorithms

Abstract

One of the greatest challenges faced by Electric Vehicle (EV) manufactures is insufficient charging stations. Estimating the aging of the battery in the electric vehicle helps the driver to predict the driving range of the vehicle. This paper proposes a battery management system that is developed to predict remaining battery charge of the Electric Vehicle. The aging of the lithium-ion (Li-Ion) battery present in the electric vehicle is predicted using different machine learning and deep learning algorithms. The parameters such as voltage, current and temperature are taken from the sensors connected to the LPC2148 ARM board and the values are given as dataset to the Long Short-Term Memory (LSTM), Decision Tree (DT), K-Nearest Neighbors (KNN), Naïve Bayes (NB) and Support Vector Machine (SVM) Algorithms. The experimental results indicate that for real-time data Naïve Bayes algorithm gave the best results in terms of metrics such as Accuracy, Precision, Recall and F1-score. Naïve Bayes produced results with the accuracy rate of 88% and used to calculate the Remaining Battery Capacity which helps predicting the aging of the lithium- ion battery.