基于鞭打试验的安全座椅神经网络优化设计

Optimal Design of Safety Seat Neural Network Based on Whipping Test

摘要: 在新版C-NCAP管理规则中，增加了颈部保护测试(鞭打测试)。汽车座椅的安全设计对乘员的安全至关重要，因此后碰撞已成为汽车被动安全的重要组成部分。本文提出了一种神经网络与遗传算法相结合的安全座椅整体优化设计方法，以防止颈部扭伤。该方法优化后的座椅在头部加速度、颈部剪切力、颈部损伤指标等方面均有改善，避免了传统安全座椅优化方法求解离散点的缺陷。与正交试验优化结果相比，NIC降低了9.0%，Nkm降低了12.5%。

Abstract: In the new version of C-NCAP management rules, the neck protection test (whipping test) has been added. The safety design of car seat is very important to the safety of passengers, so rear collision has become an important part of passive safety of cars. In this paper, an overall optimization design method of safety seat based on neural network and genetic algorithm is proposed to prevent neck sprain. The seat optimized by this method has been improved in the aspects of head acceleration, neck shear force and neck injury index, which avoids the defect of traditional safety seat optimization method in solving discrete points. Compared with the optimization results of orthogonal test, NIC is reduced by 9.0% and Nkm is reduced by 12.5%.