Optimization design for automobile ORSs based on composite deep Gaussian process regression network

doi:10.3969/j.issn.1674-8484.2025.03.002

Abstract

Abstract:

A data-driven optimization method was investigated for automobile occupant restraint systems (ORS) based on composite deep Gaussian process regression network to improve the safety performance and to develop the efficiency of the ORS. In terms of the prediction of occupant dummy injury values, an improved composite deep Gaussian process regression network was proposed as the prediction model by combining neural network architecture with Gaussian process regression. Based on the prediction results, the ORS parameter optimization was carried out by using the group-based crow search algorithm. The method’s effectiveness was verified by using engineering simulation data. The results showed that this ORS design reduces the dummy injuries by up to 30.77% with an average of 12.11% compared to the original engineering scheme. Therefore, the method can predict the injury values for multiple parts of the dummy with a high-quality ORS design.

Key words: automobile crash, occupant restraint systems (ORS), dummy injury, data-driven, composite deep Gaussian process regression network, group-based crow search algorithm

CLC Number:

U461.91

WANG Wenjie, SUN Yi, LIU Zhao, ZHU Ping. Optimization design for automobile ORSs based on composite deep Gaussian process regression network[J]. Journal of Automotive Safety and Energy, 2025, 16(3): 367-375.

Figures/Tables 9

References 22

[1]	洪亮, 葛如海. 正面碰撞中基于新型前排安全座椅的后排约束系统优化研究[J]. 汽车工程, 2015, 37(11): 1277-1283.
	HONG Liang, GE Ruhai. Research on the optimization of rear restraint system based on a new front seat in frontal crash[J]. Autom Engineering, 2015, 37(11): 1277-1283. (in Chinese)
[2]	刘鑫, 刘祥, 周振华, 等. 基于自适应代理模型的汽车乘员约束系统优化设计[J]. 汽车工程, 2020, 42(7): 887-893.
	LIU Xin, LIU Xiang, ZHOU Zhenhua, et al. Optimization design of automotive occupant restraint system based on adaptive surrogate models[J]. Autom Engineering, 2020, 42(7): 887-893. (in Chinese)
[3]	肖华, 肖森, 赵贶雨, 等. 侧面柱碰撞工况下驾驶员侧气囊优化设计[J]. 汽车技术, 2024, (04): 24-29.
	XIAO Hua, XIAO Sen, ZHAO Kuangyu, et al. Optimal design of driver’s side airbag in side pole crash[J]. Autom Tech, 2024, (04): 24-29. (in Chinese)
[4]	段大禄, 贾丽刚, 易超, 等. 小型电动汽车正面碰撞的约束系统关键设计[J]. 汽车零部件, 2023(10): 58-63.
	DUAN Dalu, JIA Ligang, YI Chao, et al. Key design of restraint system for small electric vehicle in frontal collision[J]. Autom Parts, 2023(10): 58-63. (in Chinese)
[5]	Schneider B, Kofler D, D’Addetta G, et al. Approach for machine learning based design of experiments for occupant simulation[J]. Front Future Transport, 2022, 3: 913852.
[6]	SUN Wenbo, LIU Jiacheng, HU Jingwen, et al. Adaptive restraint design for a diverse population through machine learning[J]. Front Pub Health, 2023, 11: 1202970-1202980.
[7]	李晗. 面向汽车碰撞安全数据特征的数据处理与建模方法研究[D]. 上海: 上海交通大学, 2022.
	LI Han. Research on data processing and modeling method for vehicle crash safety characterized data[D]. Shanghai: Shanghai Jiao Tong University, 2022. (in Chinese)
[8]	LI Mushi, LIU Zhao, HUANG Li, et al. Multi-fidelity data-driven optimization design framework for self-piercing riveting process parameters[J]. J Manuf Proc, 2023(99): 812-824.
[9]	FU Chongbo, DONG Huachao, WANG Peng, et al. Data-driven harris hawks constrained optimization for computationally expensive constrained problems[J]. Complex Intel Syst, 2023, 9(4): 4089-4110.
[10]	ZHANG Haobin, SAN Hongjun, CHEN Jiupeng, et al. Black eagle optimizer: a metaheuristic optimization method for solving engineering optimization problems[J]. Cluster Computing, 2024, 27(9): 12361-12393.
[11]	周亦辰. 基于自适应高斯过程回归的汽车碰撞可靠性分析与优化设计[D]. 长春: 吉林大学, 2023
	ZHOU Yichen. Reliability-based design optimization for vehicle collision based on adaptive Gaussian process regression[D]. Changchun: Jilin University, 2023. (in Chinese)
[12]	Raponi E, Bujny M, Olhofer M, et al. Kriging-assisted topology optimization of crash structures[J]. Comput Meth Appl Mech Eng, 2019, 348: 730-752.
[13]	Hay J, Lars S, Eric B, et al. Application of data-driven surrogate models for active human model response prediction and restraint system optimization[J]. Front Appl Math Statis, 2023, 9:1156785.
[14]	李启明, 张鹏飞, 喻泽成, 等. 基于各向异性混合核函数高斯过程回归的RC柱概率抗剪承载力模型[J]. 工程科学与技术, 2025, 57(01): 287-295.
	LI Qiming, ZHANG Pengfei, YU Zecheng, et al. Probability shear capacity model of rc columns based on anisotropic mixture kernel function gaussian process regression[J]. Engi Sci Tech, 2025, 57(1): 287-295. (in Chinese)
[15]	周慧灵. 基于组合核函数高斯过程回归的股指波动率预测研究[D]. 长沙: 湖南大学, 2021.
	ZHOU Huiling. Forecasting of stock index realized volatility based on Gaussian process regression with compositional kernel[D]. Changsha: Hunan University, 2021. (in Chinese)
[16]	刘迎迎, 张孝远, 刘梦楠, 等. 基于自适应最优组合核函数高斯过程回归的锂电池健康状态区间估计[J]. 储能科学与技术, 2025, 14(1): 346-357. doi: 10.19799/j.cnki.2095-4239.2024.0473
	LIU Yingying, ZHANG Xiaoyuan, LIU Mengnan, et al. Interval estimation of state of health for lithium battery based on gaussian process regression with adaptive optimal combination kernel function[J]. Energy Stor Sci Tech, 2025, 14(1): 346-357. (in Chinese)
[17]	李晗, 刘钊, 朱平. 基于SSI-PSO的汽车碰撞试验时序数据处理与分类方法[J]. 汽车安全与节能学报, 2022, 13(2): 259-268.
	LI Han, LIU Zhao, ZHU Ping. Automobile crash test time-series data processing and classification method based on SSI-PSO algorithm[J]. J Autom Safe Energy, 2022, 13(2): 259-268. (in Chinese)
[18]	Kim H Y, Won C H. Forecasting the volatility of stock price index: a hybrid model integrating LSTM with multiple GARCH-type models[J]. Expert Syst Appl, 2018, 103: 25-37.
[19]	Heidari A A, Mirjalili S, Faris H, et al. Harris hawks optimization: algorithm and applications[J]. Future Gene Compu Syst, 2019, 97: 849-872.
[20]	Askarzadeh A. A novel metaheuristic method for solving constrained engineering optimization problems: crow search algorithm[J]. Compu Struct, 2016, 169: 1-12
[21]	Arora S, Singh H, Sharma M, et al. A new hybrid algorithm based on grey wolf optimization and crow search algorithm for unconstrained function optimization and feature selection[J]. IEEE Access. 2019, 7: 26343-26361.
[22]	LIU Zhao, WANG Wenjie, SHI Guohong, et al. A modified crow search algorithm based on group strategy and adaptive mechanism[J]. Engi Opti, 2024, 56(4): 625-643.

安全带	卷收器点爆时间，T_R
	卷收器限力等级，F_L
	预紧器(PLP)点爆时间，T_P
转向管柱	压溃力，F_C
转向管柱	压溃行程，S_C
安全气囊	点爆时间，T_A
	排气孔直径，D_V
	气袋直径，D_B
	上拉带长度，L_U
	下拉带长度，L_L
	气体发生器的输出压强，p_{out, g}

安全带	卷收器点爆时间，T_R
	卷收器限力等级，F_L
	预紧器(PLP)点爆时间，T_P
转向管柱	压溃力，F_C
转向管柱	压溃行程，S_C
安全气囊	点爆时间，T_A
	排气孔直径，D_V
	气袋直径，D_B
	上拉带长度，L_U
	下拉带长度，L_L
	气体发生器的输出压强，p_{out, g}

头部	头部伤害准则(15 ms间隔)，HIC₁₅
头部	合成加速度3 ms过载量，a_{3 ms}
颈部	颈部的伸张M_yi峰值
	剪切力F_x连续过载量
	伸张力F_z连续过载量
胸部	变形峰值，D_C
胸部	粘性准则值，VC
大腿	轴向压缩力连续过载量，F_F
膝关节	位移峰值，D_K
小腿	胫骨指数值，TI
小腿	压缩力峰值， F_T

头部	头部伤害准则(15 ms间隔)，HIC₁₅
头部	合成加速度3 ms过载量，a_{3 ms}
颈部	颈部的伸张M_yi峰值
	剪切力F_x连续过载量
	伸张力F_z连续过载量
胸部	变形峰值，D_C
胸部	粘性准则值，VC
大腿	轴向压缩力连续过载量，F_F
膝关节	位移峰值，D_K
小腿	胫骨指数值，TI
小腿	压缩力峰值， F_T

部位	伤害值项目	CDGPRN	GPR-R	GPR-M	GPR-L	RS	DTR	SVR	ANN
头部	HIC₁₅	0.156 2	0.168 6	0.168 4	0.158 5	0.162 3	0.172 3	0.169 6	0.178 9
头部	a_{3 ms} / g	0.161 3	0.172 0	0.169 8	0.175 6	0.191 3	0.176 8	0.176 2	0.168 5
颈部	M_yi / Nm	0.100 3	0.107 8	0.109 8	0.109 3	0.112 0	0.127 2	0.112 8	0.111 7
	F_x / kN	0.066 7	0.070 1	0.068 2	0.073 5	0.097 7	0.082 5	0.079 5	0.075 3
	F_z / kN	0.140 5	0.165 5	0.155 7	0.168 2	0.228 5	0.196 2	0.166 4	0.150 6
胸部	D_C / mm	0.099 5	0.119 4	0.113 4	0.132 5	0.166 2	0.142 5	0.124 4	0.116 5
胸部	VC / ms^-1	0.063 2	0.083 1	0.076 0	0.078 7	0.100 5	0.080 8	0.078 5	0.061 8
大腿	F_F / kN	0.107 9	0.111 8	0.110 5	0.113 2	0.133 5	0.124 4	0.114 4	0.117 0
膝部	D_K / mm	0.053 3	0.058 2	0.064 6	0.062 0	0.060 9	0.061 6	0.069 2	0.059 8
小腿	TI	0.128 7	0.140 9	0.138 5	0.148 1	0.148 5	0.149 2	0.141 2	0.147 7
小腿	F_T / kN	0.202 2	0.218 2	0.222 4	0.235 6	0.286 5	0.256 2	0.217 5	0.210 2
各项伤害值的RMSE之和		1.279 8	1.415 6	1.397 3	1.455 2	1.687 9	1.569 7	1.449 7	1.391 7