Internal short circuit (ISC) in lithium-ion batteries is the main cause of thermal runaway accidents

and early identification of ISC faults is crucial to reduce the risk associated with fire or explosion

and traditional detection methods based on filtered capacity increment curves are susceptible to interference. Aiming at the problem of how to extract the characteristic parameters on the capacity increment curve for early internal short circuit detection

a voltage reconstruction model based on the second-order RC equivalent circuit is proposed. Firstly

the capacity increment analysis of lithium battery is carried out using the traditional filtering method

and then the voltage reconstruction model is used for comparison. And the voltage reconstruction model is subjected to anti-noise interference experiments

and the error of the voltage reconstruction is only 0.000 8 V

which lays a foundation for the early internal short circuit detection work. Finally

experiments are carried out on publicly available experimental data to investigate the effects of different charging rates and different battery types on the capacity increment curve. Based on thus

the characteristic peaks used to diagnose short-circuit faults are obtained. The experimental results show that the proposed method based on voltage reconstruction model can effectively counteract noise interference and provide reference for early micro-inner short circuit detection.