Optimal Fast Charging Method for a Large-Format Lithium-Ion Battery Based on Nonlinear Model Predictive Control and Reduced Order Electrochemical Model

Abstract

Design of an appropriate charging protocol is a challenging issue. Recent emerging charging methods have employed electrochemical models to determine charging protocols. However, these protocols are not otimized considering reduction of charging time and minimization of reaction rates of solid electrolyte phase layer (SEI) formation and lithium plating simultaneously. In this paper, we proposed an optimal fast charging method that simultaneously considers the charging time and the two aging effects in addition to lithium stripping. The optimization is performed using nonlinear model predictive control, where variables in the objective function are the charging time and ion loss caused by SEI layer and constraints are cutoff voltage and lithium ion concentrations. In addition, the charging protocol consists of not only constant currents, but also pulse discharging currents that can recover ions from the lithium metal by lithium stripping, which can further reduce the charging time by allowing increased currents. The designed charging method is verified using Battery-In-The-Loop system and compared with CC/CV charging protocols under a constant temperature. Experimental results have shown significant minimization of capacity fade and the associated extension of cycle life, and effective suppression of the lithium plating.