Accurate Monitoring of Lithium-Ion Battery Performance through Adaptive Unscented Kalman Filters

Abstract

The sources of modern energy with the best combinations of energy density, cycles, and weight include lithium-ion batteries. It covers a vast area from electric vehicles to the storage of renewable energy. Though there are some significant advantages to LiBs, estimations of the specific values of their performance parameters, State of Charge (SOC), and State of Health (SOH), are inadapted or irrelevant. They emerge from their nonlinear behavior, dynamic operating conditions, and aging effects. This paper surveys advanced estimation techniques, focusing on the Adaptive Unscented Kalman Filter, a method superior to traditional ones, such as the Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF). Its adaptability allows AUKF to improve accuracy and robustness in SOC and SOH monitoring to address the added complexities in the management of a battery. In addition, hybrid approaches combining physics-based and AI-driven models are discussed to give better estimation and to reduce modelling errors. Practical demonstrations of these approaches have been very effective in improving battery performance and safety. It also discusses the emerging trend of second-life batteries and innovative dispatch strategies that consider degradation models and real-time SOH estimation. This therefore calls for the adoption of more advanced hybrid frameworks to enhance battery efficiency, reliability, and sustainability, thus pushing the frontiers of energy storage technologies.