A multi-objective optimization method based on an adaptive combined surrogate model for the front seat of a passenger car

Abstract

The optimized design of automotive seats is critical to improving passenger safety during high-speed collisions. To address the issues of efficiency and cost in optimizing automotive seats, a detailed method for optimization design is proposed. Firstly, based on experimental data, two types of finite element models for dynamic crash working conditions of automotive seats are established, and their accuracy is verified. Secondly, the design variables are screened through comprehensive contribution analysis, and fifteen thickness variables and seven material variables are ultimately selected. Subsequently, the prediction model of the input variables and output response of the automotive seat is constructed using the adaptive combined surrogate model, and the non-dominated sorting genetic algorithm III is employed to determine the pareto solution set. Finally, the optimal combination weights derived from the fuzzy analytic hierarchy process and the coefficient of variation method are utilized to assign weights to each performance index of the automotive seat. The relative entropy-based technique for order preference by similarity to an ideal solution method is employed to evaluate the pareto solution set and determine the optimal trade-off solution. In addition, the effectiveness of the proposed optimization design method is verified by comparing the baseline design, simulation analysis, and optimization design methods. The optimization results indicate that, while ensuring safety performance, the material cost and mass of the seat skeleton were reduced by 9.01% and 10.42%, respectively. The multi-objective optimization strategy presented in this paper demonstrates strong optimization capability and efficiency while providing a reliable reference for automotive seat design.