Online adaptive anode potential-controlled fast charging of lithium-ion cells using a validated electrochemical model-based virtual reference electrode

Abstract

Optimal utilization of the safe operating area of lithium-ion battery cells is fundamental to achieving peak performance. Pushing the currents close to the intrinsic electrochemical cell limits during fast charging of battery electric vehicles reduces charging times noticeably and improves customer satisfaction. With electrochemical model-based state observers embedded in the battery management system, predicting the critical anode potential is feasible. This allows the charging process to be adjusted accordingly for effective prevention of lithium plating. However, validating the model, especially the internal potentials, is particularly challenging, and the real-time capability is often doubted. This work addresses the parameterization and experimental validation of a full-order electrochemical model utilizing cells instrumented with micro-reference electrodes. Special emphasis is put on validating the estimate of the anode surface potential, which enables the model to act as a virtual reference electrode. For the first time, it is shown how this can be efficiently implemented on a next-generation automotive microcontroller, in combination with a charging current controller, for online adaptive anode potential-controlled fast charging. Execution times well below 4 ms prove real-time capability, and reference electrodes validate good agreement between model estimate and measurements in closed-loop charging current control.