Optimized Support Vector Regression model for accurate state of charge estimation of Li-Ion Phosphate batteries for Electric Vehicles

The proposed work approaches machine learning based Lithium Iron-phosphate (LIP) battery state of charge (SoC) forecasting method for electric vehicles (EVs). The SoC must be accurately estimated to operate safely and steadily for an LIP battery. It is challenging for the user to determine the battery SoC during the charging segment because of the random nature of their charging process. One of the fundamental components of artificial intelligence, machine learning (ML) is rapidly transforming a wide range of fields with its capacity to learn from given data and resolve complex problems. A Grey Wolf optimization (GWO) algorithm is used to modify the most pertinent hyperparameters of the SVR model to enhance the accuracy of the predictions. The database used in the study was gathered from a laboratory setup and includes information on temperature, voltage, current, average voltage, and, average current. Initially, a preprocessing method for data normalization is utilized to improve the data quality and forecast accuracy. It is demonstrated that the accuracy of the suggested model for estimating the SoC of the LIP battery shows less than % MSE between the measured and expected states of charge.