考虑驾驶风格的换道预警方法.pdf

Methods of Lane Change Warning Considering the Driving Style A Dissertation Submitted for the Degree of Master Candidate：：Mao Jin Supervisor：：Prof. Fu Rui Chang’an University, Xi’an, China I 摘 要 换道过程是常见的驾驶行为之一，也是一种容易引发交通冲突的过程，因此研究人员提出了使用换道预警系统来提高换道安全性 目前换道预警系统通常采用固定参数进行预警， 但由于驾驶人的驾驶风格存在差异， 因此部分驾驶人对此类系统存在抵触情绪，影响了换道预警系统的推广使用针对此问题，本文通过实际道路换道实验，采集驾驶人的实际换道数据对驾驶风格进行分类， 并对不同风格的驾驶人采取个性化的换道预警参数，降低换道预警系统对驾驶人带来的干扰 首先，根据实际道路驾驶实验数据，使用多因素模糊数学分类法对驾驶人的驾驶风格进行分类通过使用跟车时距、最小跟车时距、换道时距、超速频次、换道过程方向盘最大转角与车速拟合结果等 5 个参数，将驾驶人分为冒进型、比较冒进型、比较谨慎型和谨慎型四类。

其次，通过建立车道变换过程模型，理论分析影响安全车道变换的因素，提出车道变换的预警指标 最后，通过分析不同类型驾驶人的实际换道数据值，并参考理论经验值，得出不同类型驾驶人的换道预警指标值利用新的被试驾驶人对驾驶风格分类模型进行验证，结果表明，本文所提出的驾驶风格模型能实现对不同风格驾驶人的换道预警功能 本研究得到了国家自然科学基金项目（51178053）和国家科技支撑计划项目（2009BAG13A05）的资助 关键词：换道预警，模糊数学，驾驶风格，预警参数 II Abstract As lane change is one of the common driving behavior and also a kind of process which is easy to cause traffic conflict, researchers proposed to use the lane change warning system to improve driving safety. Present lane change warning system is usually a fixed warning parameter which can not adapt to the differences of driver’s driving style and was not accepted by drivers, so the promotion and application of this system was affected. For this problem, this paper collected the lane change data of drivers by actual road test. Besides the driving style was classified and personalized lane change warning parameters were set up according to different driving style which can reduce interference with drivers caused by the system. Firstly, according to the data of actual road test, the driver's driving style was classified using of multi-factor fuzzy mathematical taxonomy. Driving people were divided into the four classes aggressive, relatively aggressive, relatively cautious and cautious through 5 parameters. The 5 parameters included time-distance between followed vehicles, the smallest time-distance between followed vehicles, lane change time-distance, the frequency of over-speeding and the fitting results of maximum steering wheel angle with vehicle velocity in lane change process. Secondly, factors that affect the safety of lane change process were theoretically analyzed through the establishment of the lane change process model, and the lane change warning indicators were putted forward. Finally, the lane change warning indicator values which adapted to different types of driver was got by analyzing the actual lane change data of different types of drivers and referring to a theory and experience value. The model of driving style classification and the model of personalized lane change warning was verified using of new sample, the results show that the model of driving style classification can realize warning function for different driving style in lane change process. The research was sponsored by National Natural Science Foundation (51178053) and National Key Technology R fuzzy mathematics; driving style; warning parameter III 目目 录录 第一章第一章 绪论绪论 .............................................................................................................................. 1 1.1 研究背景及意义 ........................................................................................................... 1 1.2 国内外文献回顾 ........................................................................................................... 2 1.2.1 换道模型 ............................................................................................................. 2 1.2.2 换道预警 ............................................................................................................. 4 1.3 本文主要研究内容 ....................................................................................................... 6 第二章第二章 基于模糊数学的驾驶人驾驶风格分类基于模糊数学的驾驶人驾驶风格分类 ...................................................................... 8 2.1 实验介绍 ....................................................................................................................... 8 2.1.1 实验车辆 ............................................................................................................. 8 2.1.2 实验仪器 ............................................................................................................. 8 2.1.3 实验路段 ............................................................................................................. 9 2.1.4 驾驶人选择 ....................................................................................................... 10 2.2 基于模糊数学的模型建立 ......................................................................................... 11 2.2.1 数学模型的表述 ............................................................................................... 11 2.2.2 隶属函数的确定 ............................................................................................... 12 2.2.3 评价指标的选择 ............................................................................................... 13 2.2.4 权重的确定 .....................................................................................................。