Multi-Objective Optimization of Multi-Stage Constant Voltage Charging Strategy With Stepwise Ramp-Up for Lithium-Ion Batteries

This study proposes a multi-stage constant voltage (MSCV) fast-charging strategy for lithium-ion batteries, in which state of charge (SOC) is used as the transition criterion. To balance computational efficiency with physical fidelity, a hierarchical modeling framework is adopted: a second-order Thevenin equivalent circuit model (ECM) drives the iterative optimization, while a coupled electrochemical pseudo-two-dimensional (P2D) and thermal model validates the internal safety of candidate profiles. Nine MSCV design parameters—five stage voltages and four SOC transition points—are optimized simultaneously to balance charging time and energy loss. Unlike single-algorithm studies, this work conducts a comprehensive benchmarking of various multi-objective evolutionary algorithms (MOEAs), identifying NSGA-II and MOGEO as the most robust solvers in terms of convergence and diversity. The analysis reveals pronounced current spikes at stage transitions, motivating the proposal of a stepwise voltage ramp-up mechanism. This mechanism bridges the gap between aggressive theoretical optimization and practical hardware constraints by effectively dampening transient surges. The optimized MSCV schemes are validated experimentally on an INR21700 M50LT cell. Results show that in the shortest-charging-time case, the total time is reduced by approximately 6.4% compared to constant current–constant voltage (CC–CV), with a 1.8% decrease in average temperature and peak temperatures maintained within safe limits. These outcomes demonstrate that the proposed strategy significantly improves fast-charging efficiency and thermal management while ensuring physical plausibility.