基于ANFIS的小重叠碰撞安全气囊算法研究  

A Study on Airbag Algorithm for Small-overlap Crash Based on ANFIS

作　　者：白中浩[1] 张林伟[1] 白芳华 颜长征[2] 覃祯员[2] 张永春[2] 

机构地区：[1]湖南大学,汽车车身先进设计制造国家重点实验室,长沙410082 [2]重庆车辆检测研究院,国家客车质量监督检验中心,重庆401122

出　　处：《汽车工程》2017年第5期524-529,523,共7页Automotive Engineering

基　　金：国家自然科学基金(51475153)资助

摘　　要：为提高小重叠碰撞事故中安全气囊对乘员的保护作用,分析了小重叠碰撞事故类型并获得识别小重叠碰撞的3个特征参数:速度变化量、加速度曲线长度比值和汽车主要受力方向。基于特征参数变量,结合自适应神经模糊推理系统开发了新型二级模糊安全气囊控制算法,来识别小重叠碰撞类型,并根据碰撞类型展开安全气囊。通过训练合适的模糊隶属度函数和模糊规则,该算法的性能得以提高。最后,经台车试验对新型安全气囊算法进行了验证。结果表明:在发生小重叠碰撞事故时,基于ANFIS的新型安全气囊控制算法能准确识别小重叠碰撞类型,及时展开前安全气囊和侧气帘,改善了其对乘员的保护作用。In order to enhance the protection effects of safety airbag for occupant in small-overlap crash, the accident types of small-overlap crash are analyzed with three characteristic parameters obtained for identifying small-overlap crash, i. e. speed change,acceleration curve length ratio and the principal direction of force. Based on characteristic parameters,combined with adaptive neural fuzzy inference system （ ANFIS） ,a novel two stage fuzzy airbag control algorithm is developed to identify the type of small-overlap crash, according which the safety airbag is deployed. The performance of the algorithm can be enhanced by training appropriate fuzzy membership functions and fuzzy rules. Finally,the new airbag algorithm is validated by sled test. The results show that in the event of small- o-verlap crashes,the new airbag algorithm based on ANFIS can accurately identify the type of small-overlap crash and timely deploy front airbag and side air- curtain,improving their protection effects for occupant.

关 键 词：乘员保护 小重叠碰撞 自适应神经模糊推理系统 安全气嚢算法 

分 类 号：U491.61[交通运输工程—交通运输规划与管理]