Lithium-ion battery, electrochemical model, stability analysis, time-space grid optimization, swarm intelligence algorithm, Electrical engineering. Electronics. Nuclear engineering, TK1-9971

Lithium-ion batteries have been widely adopted as power sources in automotive and other fields. Balancing voltage prediction accuracy and simulation time is critical for the performance of battery management systems (BMS). This study proposes a time space grid optimization method based on the single particle model with electrolyte dynamics (SPMe) to enhance the computational efficiency and accuracy of electrochemical models while achieving a trade-off between voltage estimation precision and computational time. A second-order finite difference algorithm and Fourier stability analysis were used to determine the convergence constraints for SPMe model concentration calculations. A multi-objective time space grid optimization model was developed and solved using a multi-objective artificial bee colony algorithm. The effectiveness of the proposed method was validated by comparing simulation results under different scenarios. The model provides optimal time space grid configurations for various current profiles, including 1C-rate constant-current charging, pulse current (PC), hybrid current (HC), and urban dynamometer driving schedule (UDDS) current profiles. The SPMe model achieved a voltage estimation root mean square error (RMSE) of less than 5 mV, while maintaining a simulation time of less than 10 ms per second. This work provides a new idea for voltage estimation accuracy and calculation time balance of existing battery management systems under complex operating conditions.