Lithium-ion battery electro-thermal modelling and internal states co-estimation for electric vehicles

Battery modelling, temperature monitoring and accurate estimation of capacity and state of charge (SOC) are fundamental functions of the battery management system (BMS) for ensuring the safety and reliability of lithium-ion batteries (LIBs). Most studies focus on estimating one or two of these states while ignoring the influence of temperature and ageing on the battery's performance. Neglecting the coupling effects between the battery capacity, SOC and temperature will hinder the accuracy of their estimates. This study uses an enhanced electrothermal battery model (EETM) to build a co-estimation scheme of the battery capacity, SOC, core and surface temperature. Regular updates of the battery capacity help track the battery's health and enhance the precision of the SOC estimation results. The model monitors the battery core and surface temperature inside a 1s3p 18650 battery pack to prevent thermal runaway. The battery model and the co-estimation system is validated using COMSOL Multiphysics (R) simulation software. Results show that real-time updates of the battery operating temperatures, capacity, and SOC yielded good results and improved the estimators' accuracy. Using the urban dynamometer driving schedule (UDDS) test, the proposed approach was able to reduce the estimation errors to +/- 9 mV for the battery model output voltage, 0.5% for the SOC and 0.5 K for the core temperature.