Comprehensive Equivalent Circuit Based Modeling and Model Based Management of Aged Lithium ion Batteries.

摘要

Energy storage is one of society's grand challenges for the 21st century. Lithium ion batteries (LIBs) are widely used in mobile devices, transportation, and stationary energy storages due to lowering cost combined with excellent power/energy density as well as cycle durability. The need for a battery management system (BMS) arises from a demand to improve cycle life, assure safety, and optimize the full pack performance. In this work, we proposed a model based battery on-line state of charge (SoC) and state of health (SoH) estimator for LIBs. The estimator incorporates a comprehensive Equivalent Circuit Model (ECM) as reference, an Extended Kalman Filter (EKF) as state observer, a Recursive Least Square (RLS) algorithm as parameter identifier, and Parameter Varying Approach (PVA) based optimization algorithms for the parameter function regressions. The developed adaptive estimator was applied to a 10kW smart grid energy storage application using retired electric vehicle batteries. The estimator exhibits a high numerical efficiency as well as an excellent accuracy in estimating SoC and SoH. The estimator also provides a novel method to optimize the correlation between battery open circuit voltage (OCV) and SoC, which further improves states estimation accuracy.