Abstract: An electro-thermal-aging coupling multi-state joint-estimation model of lithium-ion batteries was built based on the first-order RC equivalent circuit model， the thermal network method and the aging model. Particle swarm optimization(PSO) algorithm was adopted to establish a self-adaptive multistage constant current(SMCC)charging strategy whose stages were adaptive to the objective function. By constructing the Pareto frontier of charging time and battery life， three charging strategies， including the minimum-time charge， the minimum-aging charge and the balanced charge， were obtained. Then they were compared with the CC-CV charges. The results show that the minimum-time charge is highly consistent with the 2C CC-CV. The aging losses of the minimum-aging charge and the 0.1C CC-CV are very small， but the former reduces charging time by 61.7%. Compared with the minimum-aging charge， the charging time of the balanced charge is reduced by 71.19% at the expense of 0.06%SOH. Compared with the 0.5C CC-CV charge， the charging time of the balanced charge is reduced by 44.9%.

Key words: lithium-ion batteries, electro-thermal-aging coupling model, SMCC(self-adaptive multistage constant current), particle swarm optimization(PSO) algorithm, Pareto frontier