Multi-objective optimization design of a battery pack system for crashworthiness and lightweight based on a submodel and hybrid weighting method

The research focuses on optimizing battery pack systems (BPSs) for crashworthiness and lightweight design in electric vehicles. Traditional methods have high computing costs and do not account for the battery cell's failure criterion. This study uses a submodel and hybrid weighting for multi-objective design. It starts with the finite element model of a single cell, validates it and then builds a BPS model with a submodel for key areas. The optimal Latin hypercube sampling technique is used to sample the design space, and a kriging surrogate model and multi-objective particle swarm algorithm are used to find the Pareto solution. The compromise solution is selected using the analytic hierarchy process–Entropy–VIKOR method, considering the maximum principal strain of the cell. This approach cuts the computation time by 56.2% and offers a practical solution for BPS design by utilizing the load-bearing ability of the BPS.