Low temperatures seriously affect the performance of lithium-ion batteries. This study proposes a non-destructive low-temperature bidirectional pulse current (BPC) heating method. Different from existing heating approaches, this method not only optimizes heating frequency and amplitude but also considers the optimization of the charge/discharge pulse duration ratio. To optimize the BPC heating strategy, a precise electro-thermal coupled model is established, and a neural network is employed to delineate the relationship among model parameters, temperature, and state of charge (SOC). Additionally, the interplay between the impedance of the graphite anode and that of the full cell is analyzed by constructing a three-electrode battery. Then, a novel full-cell-oriented lithium plating criterion is introduced. Finally, based on the constructed electro-thermal coupled model, lithium plating criterion, and terminal voltage constraint, a novel non-destructive BPC heating method is proposed. The results show a significant improvement in heating efficiency compared to conventional BPC heating. Especially for high SOCs, the heating power is increased at least 8 times. When the battery SOC is below 40 %, the average heating rate from −10 °C to 10 °C is 11.28 °C/min. Even at 90 % SOC, the heating rate remains at 2.88 °C/min. Furthermore, the capacity and impedance of a battery at 50 % SOC exhibit no significant changes after 60 heating cycles using the optimal BPC heating strategy at 100 Hz. These findings show that the optimized method proposed in this study has high heating efficiency and no damage to the battery.
Read full abstract