Maximum Available Capacity and Energy Estimation Based on Support Vector Machine Regression for Lithium-ion Battery

The practical application of electric vehicle needs an accurate and robust battery management system to monitor the battery state in real-time. The maximum available capacity (MAC) and maximum available energy (MAE) need to be derived before calculating state of charge and state of energy. However, the estimation of these two parameters is a difficult task due to the complicated and comprehensive influences of temperature, aging level and discharge rate. In this paper a data-driven algorithm, least squares support vector machine, is implemented to estimate the MAC and MAE, and the influences of temperature and degradation are taken into consideration. Meanwhile, a current correction term is proposed to compensate the effect of current rate. The experimental results verify the proposed methods have excellent estimation accuracy for LiFePO4 battery.