Research on Battery State Estimation Based on Machine Learning

Abstract

In the context of the big era, the demand for battery technology in many fields is growing rapidly, and the accuracy of the accompanying battery management system (BMS) directly affects vehicle range prediction, charging strategy optimization, and safety warning. However, batteries exhibit significant nonlinear characteristics in actual operation, and traditional equivalent circuit models are difficult to meet the state estimation requirements under high-power conditions. Machine learning techniques, with their powerful nonlinear fitting capabilities and data-driven characteristics, offer new ideas for battery state estimation. This study conducts battery state estimation based on deep learning, which is of great theoretical significance and engineering application value for improving the performance of battery management systems and promoting the development of the new energy industry.