Multiparameter warning of lithium-ion battery overcharge-thermal runaway
Jianfeng Wang,Bowei Chen,Yuhan Li,Ting Hu,Fen Liu,Mengyu Shi,Xutong Ren,Yongkai Jia,Weihua Li
DOI: https://doi.org/10.1016/j.est.2023.110088
IF: 9.4
2023-12-22
Journal of Energy Storage
Abstract:The rapid development of new energy vehicles has drawn widespread attention to battery safety. Overcharging, as an important source of thermal runaway, may occur instantaneously without obvious signs, and any corresponding fire will be difficult to extinguish. This study is an investigation of overcharging thermal runaway and thermal runaway warnings for lithium-ion batteries. A stress-type early warning system is proposed, which has faster response time and more distinctive characteristics compared with other parameters. Through the association rule mining method, a multi-parameter coupled thermal runaway early warning strategy based on voltage, temperature, and pressure parameters was designed. A hierarchical early warning model including feature extraction, data processing and early warning evaluation modules was established. On this basis, a remaining time prediction module was added to achieve an alarm escape time of up to 474 s and shortest of 65 s, meeting safety standards. In the thermal runaway experiment at 705.2 °C, the early warning level system was triggered respectively. The maximum battery temperatures were 28.4 °C, 41.5°Cand 60.3 °C. The escape time errors were 16.56 s, 11.52 s, and 11.88 s respectively, all within 20 s for each level. Corresponding to different experimental results, the significant level classification simultaneously verifies the accuracy and effectiveness of the classification strategy.
energy & fuels