State of Charge Estimation for Electric Vehicle Battery Using Support Vector Machines Technique

State-of-charge (SOC) estimation is considered as the fundamental challenge for battery management system (BMS). the available SOC estimation methods face difficulty for accurate SOC estimation with different conditions such as (ambient temperature, the speed of wind or the humidity, etc). In this study, the training data and the testing data required in the estimation operation are collected using a real battery pack then followed by normalization. Then, lab experiment with MATLAB simulation applied to validate the results and get a powerful source of data generator. Then, the parameters in the support vector machine for regression model tuned to get the best estimated for SOC. Finally, simulation experiments have been done using python script to train the model and test a new data set. The results show the efficiency of using support vector machine for regression in estimating the battery SOC.