Performance reliability analysis and optimization of lithium-ion battery packs based on multiphysics simulation and response surface methodology

Abstract Reliability optimization has always been an important topic in the application of lithium-ion batteries in electric vehicles. To optimize the redundancy and layout design of battery packs accurately and efficiently, a novel reliability optimization method based on a multiphysics coupling simulation and a response surface methodology is proposed. An electrochemical-thermal-fluid dynamics model, a temperature-dependent stochastic degradation model and a multistate performance reliability model are developed. A response surface method with the Box-Behnken design method is applied to reduce the number of simulation trials. Then, the redundancy scheme and two layouts of a battery pack are optimized, followed by a sensitivity analysis of the design parameters. The results show that a large spacing in the direction of air flow contributes to improving system reliability. In this case, for the layout of a cross arrangement with equal spacing, the optimal reliability design scheme of a 5 × 5 parallel series with optimal design parameters can improve the cycle life from approximately 1989 to 2933 when taking 90% system reliability as the criterion. The collaborative optimization of redundancy and layout is of great importance to extend the service life and improve the system reliability of battery packs.