In this paper, the safety performance model of cylindrical lithium-ion batteries, which is based on a second-order oscillation feature that is subjected to mechanical abuse is proposed via a discrete Fourier transformation of experimental data. Combined with the safety performance model and Crushable-Foam material model, the simulation results show that the average predictive error of the force-displacement and failure displacement is less than 8.8% and 4.0%, respectively. A battery fracture angle model is proposed, the fracture angle is only related to the load property, and the state-of-charge has minimal influence under the same load conditions. Furthermore, the failure prediction model of the soft short-circuit and the failure warning verification system are proposed. The experimental results show that the failure prediction model can warn the soft short-circuit in time with a certain safety margin. With further research, the safety performance model and the failure prediction model, which is based on mechanical penetration displacement, can be combined with other failure signals to provide more reliable information about battery safety. The failure warning device has the advantages of low cost and simple structure.

本文通过对实验数据进行离散傅立叶变换，提出了一种基于二次振荡特征的圆柱型锂离子电池安全性能模型，该模型受到机械滥用。结合安全性能模型和可破碎泡沫材料模型，仿真结果表明，力位移和破坏位移的平均预测误差分别小于8.8％和4.0％。提出了电池断裂角模型，该断裂角仅与负荷特性有关，在相同负荷条件下，充电状态的影响最小。此外，提出了软短路故障的预测模型和故障预警验证系统。实验结果表明，该故障预测模型可以在一定安全裕度的情况下及时警告软短路。通过进一步的研究，可以将基于机械穿透位移的安全性能模型和故障预测模型与其他故障信号相结合，以提供有关电池安全性的更可靠信息。该故障预警装置成本低廉，结构简单。