Although digital twin high-fidelity models have demonstrated accuracy and effectiveness in battery system condition monitoring, performance prediction, and other fields, they have not yet been deeply integrated into the static–dynamic coupled structural optimization of battery enclosures. Taking the digital twin as a breakthrough engine, this study first conducts risk assessment on the enclosure under multigradient loading conditions based on a high-fidelity digital twin model iteratively corrected by the multiobjective particle swarm optimization (MOPSO) algorithm (maximum relative error not exceeding 8%). Then, it optimizes the topography of the upper cover with the goal of maximizing the bending and torsional modal frequencies, applies opposing compressive forces along the X- and Y-axes to the bottom plate, and reversely derives the optimal arrangement of reinforcing beams based on the distribution characteristics of reinforcing ribs output from the optimization. Finally, to fully tap the design potential, multiobjective optimization on the newly designed enclosure is conducted in combination with the sequential quadratic programming (SQP) algorithm. The results showed that after optimization, the total mass of the battery enclosure was reduced by 10.88%, its crush resistance under a 100-kN load was improved by 24.5%, the first-order bending mode was increased by 12.3%, and the second-order torsional mode was increased by 13.9%, providing a more targeted and practical technical path for the design of battery enclosures for electric vehicles.
Li et al. (Fri,) studied this question.