A novel optimization strategy combining hybrid surrogate modeling for multi-objective design of a rear seat backrest frame in a passenger car

To address the challenges associated with complex weight determination in multi-criteria decision-making and the limited predictive capability of single surrogate models for highly nonlinear crash responses in automotive seat optimization, this study proposes an integrated multi-objective global optimization (MOGO) framework incorporating a Modified Preference Selection Index (MPSI) method and a hybrid surrogate model. First, a high-fidelity finite element model of a rear-seat luggage compartment impact scenario was developed and validated against experimental results. Key design variables were then identified based on engineering constraints, and representative sample points were generated using a design of experiments approach. Subsequently, a hybrid surrogate model was constructed to establish the relationship between design variables and performance responses. By coupling the surrogate model with the Non-dominated Sorting Genetic Algorithm II (NSGA-II), a set of Pareto-optimal solutions was obtained. The MPSI method was further employed to rank the Pareto solutions and identify the optimal compromise design. To verify the effectiveness of the proposed framework, rear-seat luggage impact, headrest static strength, and modal tests were conducted on the optimized seat structure. The numerical simulation results demonstrated that the optimized design satisfied all regulatory requirements while achieving improved structural performance. The proposed optimization framework provides an effective and computationally efficient approach for solving complex multi-objective lightweight design problems and offers valuable guidance for the lightweight development of automotive structural components.