摘要
基于MADYMO车体56km/h100%正面刚性碰撞,以加权综合损伤值(WIC)为目标对乘员约束系统进行优化.选取坐姿(座椅)、安全带、安全气囊、转向柱的8个参数作为设计变量进行最优拉丁超立方设计(optimal Latinhypercube design,OptLHD)并完成135组仿真.考虑各设计变量交互效应以及非线性影响,对设计变量和仿真结果进行椭圆基函数(EBF)神经网络建模,分析单一设计变量和两变量交互取值对输出响应的影响,并在此基础上采用多岛遗传算法(MIGA)确定正面碰撞约束系统的最佳参数匹配.将最佳参数导入MADYMO进行仿真验证,结果表明误差合理且WIC下降2754%,约束系统的保护性能得到了全面提升.
Based on the MADYMO 56 km/h 100%frontal rigid impact,the occupant restraint system was optimized with WIC as the target.Eight parameters of sitting posture(seat),safety belt,airbag and steering column are selected as design variables for Opt LHD experimental design and 135 groups of simulations are completed.Considering the interaction and nonlinearity effects of design variables,the EBF neural network is used to model the design variables and simulation results,and the influence of single design variable and the interaction of two variables on the output response is analyzed.On this basis,MIGA is used to determine the best parameter matching of frontal impact constraint system.The parameters are imported into MADYMO for simulation and the error is reasonable and the WIC is reduced by 27.54%.The protection performance of the constraint system is improved comprehensively.
作者
袁庆磊
赵翼飞
邹魁
马磊
YUAN Qinglei;ZHAO Yifei;ZOU Kui;MA Lei(School of Mechanical Power and Engineering,Shenyang University of Chemical Technology,Shenyang 110142,China)
出处
《沈阳化工大学学报》
2025年第2期193-202,共10页
Journal of Shenyang University of Chemical Technology
