基于小重叠碰撞的自适应安全气囊控制方法及实现

发布时间：2018-07-08 17:33

本文选题：自适应安全气囊 + 控制算法　；参考：《湖南大学》2015年硕士论文

【摘要】：安全气囊是碰撞事故中挽救乘员生命的重要辅助装置,当汽车发生正面碰撞时,与安全带配合使用可有效降低乘员的伤亡率。为了提高安全气囊在发生小重叠碰撞事故时对乘员的保护作用,使其更加适应复杂的现实碰撞,本文提出了一种新型自适应安全气囊控制方法,并通过台车试验验证了该控制系统的可靠性。首先,分析了小重叠碰撞事故类型特征和传统安全气囊控制算法的优缺点,提出基于自适应模糊神经推理系统(Adaptive Network-based Fuzzy Inference System,ANFIS)的二级模糊安全气囊碰撞类型识别算法策略。根据经过验证的碰撞仿真模型分析得到小重叠碰撞事故的三种特征参数:速度变化量、汽车主要受力方向和加速度长度比值。以此三种特征参数作为二级模糊气囊碰撞类型检测算法的模糊输入,并通过使用ANFIS训练合适的模糊隶属度函数和模糊规则来提高该算法的性能。针对开发的二级模糊算法,分别选择了速度长度比值和汽车主要受力方向单个特征参数作为算法的模糊输入变量,与本文提出的算法做对比,分析了以不同特征参数为输入变量的二级模糊算法在识别碰撞类型方面的优缺点,初步验证了本文提出的碰撞类型识别算法的实时性。基于本文提出的碰撞类型识别算法、图像识别并融合体压分布检测乘员识别分类算法,设计了一款自适应安全气囊控制系统,当碰撞事故被识别为小重叠碰撞事故类型,并且超过安全气囊触发阈值后,在最佳时间同时展开前排乘员气囊和侧气帘,合理匹配安全气囊的形状,从而达到对不同体征乘员不同碰撞事故类型都起到最佳保护的目的。最后,综合考虑本文提出的自适应安全气囊算法特征与本实验室现有设备,进行了台车验证试验。在验证试验过程中,安全气囊均能按照预期准确识别碰撞类型和起爆条件,给乘员提供最佳保护,验证了本文提出的自适应安全气囊算法的有效性和可靠性。
[Abstract]:Airbag is an important auxiliary device to save the occupant's life in the collision accident. When the car has the frontal impact, the use with the safety belt can effectively reduce the casualty rate of the occupant. In order to improve the protective effect of airbag on occupants in the event of small overlap collision, a new adaptive airbag control method is proposed in this paper. The reliability of the control system is verified by the bench car test. Firstly, the characteristics of small overlap collision accidents and the advantages and disadvantages of traditional airbag control algorithms are analyzed, and a two-level fuzzy airbag collision type identification algorithm based on Adaptive Network-based Fuzzy reference system (ANFIS) is proposed. According to the analysis of the verified collision simulation model, three characteristic parameters of the small overlap collision accident are obtained: the variation of velocity, the main force direction of the vehicle and the ratio of acceleration length. The three characteristic parameters are used as the fuzzy input of the two-stage fuzzy airbag collision detection algorithm, and the performance of the algorithm is improved by using ANFIS to train appropriate fuzzy membership function and fuzzy rules. According to the developed two-level fuzzy algorithm, a single characteristic parameter of the speed length ratio and the main bearing direction of the vehicle is selected as the fuzzy input variable of the algorithm, which is compared with the algorithm proposed in this paper. The advantages and disadvantages of the two-level fuzzy algorithm with different characteristic parameters as input variables for collision type recognition are analyzed and the real-time performance of the collision type recognition algorithm proposed in this paper is preliminarily verified. An adaptive airbag control system is designed based on the collision type recognition algorithm proposed in this paper, image recognition and body pressure distribution detection crew identification and classification algorithm. When the collision accident is identified as a small overlap collision accident type, a self-adaptive airbag control system is designed. After exceeding the threshold of airbag trigger, the front passenger airbag and side air curtain are expanded simultaneously at the best time to match the shape of airbag reasonably, so as to achieve the purpose of best protecting the occupants with different signs and different types of collision accidents. Finally, considering the characteristics of the adaptive airbag algorithm and the existing equipment in our laboratory, the vehicle verification test is carried out. During the verification test, the airbag can accurately identify the collision type and detonation condition according to the expectation, and provide the best protection for the occupants. The validity and reliability of the adaptive airbag algorithm proposed in this paper are verified.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U467.14;U491.61

【参考文献】