Collaborative optimization of body-in-white and battery pack structure based on a two-stage hybrid surrogate model

To address the collaborative optimization needs of lightweight design and structural reliability in new energy vehicles, this study establishes an integrated multi-objective optimization framework for vehicle body-battery pack systems. Based on a specific vehicle model, a finite element model encompassing the modal characteristics of the body-in-white, static stiffness, and battery pack constraint modes was developed, along with a static simulation system that incorporates five working conditions. This study employs a two-stage optimization strategy. Initially, the Particle Swarm Optimization algorithm is utilized to perform a global optimization of the initial weights and thresholds of the BP neural network, thereby enhancing the predictive accuracy of structural performance parameters for the vehicle body and battery pack. Subsequently, the trained predictive model is integrated with an improved NSGA-II optimization algorithm to achieve a synergistic optimization of the structural performance of the vehicle body and battery pack. After optimization, the bending stiffness of the white body is increased by 5.57%–9287.9 N/mm, the first-order modal frequency of the battery pack is enhanced by 28.38%–29.45 Hz, and the weight of the white body is reduced by 2 kg (0.22%). The proposed optimization scheme overcomes the limitations of traditional experimental methods in terms of operational complexity and time consumption, achieving a balance in structural performance between the white body and battery pack under the premise of lightweight design, thereby providing an effective solution for the multi-system collaborative design of new energy vehicles.