Engine hood is one of the important parts of the vehicles, which has influences on the lightweight, structural safety, pedestrian protection, and aesthetics. The optimization design of engine hood is a high-dimensional, multi-objective, and mixed-variable optimization problem. In order to reduce the physical test investment in the development and improve the efficiency of optimization, this article proposes a data-driven method for optimal hood design. A newly proposed single-objective optimization algorithm is improved by several strategies for multi-objective constrained problem with mixed variables. Then the hood is optimized through the specially designed machine learning model. Finally, both the hood’s weight and pedestrian injury are reduced while maintaining structural stiffness and frequency in the desired range. The comparative study and final hood optimization results prove the effectiveness of the proposed method.

发动机罩是车辆的重要组成部分之一，对轻量化，结构安全，行人保护和美观都有影响。发动机罩的优化设计是一个高维，多目标，混合变量的优化问题。为了减少开发过程中的物理测试投资并提高优化效率，本文提出了一种数据驱动的优化引擎盖设计方法。针对混合变量的多目标约束问题，通过几种策略对新提出的单目标优化算法进行了改进。然后，通过特殊设计的机器学习模型对引擎盖进行优化。最后，在保持结构刚度和频率在所需范围内的同时，减轻了发动机罩的重量和行人受伤。