基于NAIS的AEB系统路口测试场景研究

为建立适合中国道路的自动紧急制动系统路口测试场景。对国家车辆事故深度调查体系（NAIS）数据进行筛选得到582起路口车-车事故案例，运用聚类方法对事故数据进行分析，并根据事故样本数和事故伤亡程度对聚类结果进行典型参数提取，得到17类典型路口车车事故场景。在典型事故场景的基础上分析并加入2车最高碰撞车速、碰撞角度和碰撞类型3个关键参数，使最终得到的测试场景比已有的研究更能准确地反映实际的事故统计情况。      

汽车VS二轮车十字路口事故场景研究

为研究十字路口发生的汽车与二轮车事故场景,基于NAIS事故数据库中发生在十字路口的汽车与二轮车事故数据开展研究。选择描述十字路口事故关键参数,依据事故发生时环境照明、十字路口信号灯、汽车所在道路车道数、道路限速、汽车种类、汽车速度、汽车运动方式、第一碰撞点、二轮车种类、二轮车运动方式等参数,对事故数据进行聚类分析,得到5类主要危险场景。基于发生频率最大和骑车人死亡风险最高的场景具有危险场景代表性,最终得到2类十字路口汽车与二轮车的典型事故场景。为十字路口汽车与二轮车的冲突消解、避撞策略研究提供参考。     

Risk assessment based on driving behavior for preventing collisions with pedestrians when making across-traffic turns at intersections

Traffic accident statistics in Japan show the necessity of preventing vehicle-on-pedestrian accidents. If the risk of a vehicle colliding with pedestrians could be evaluated in advance, driver-assistance systems would be able to support drivers to avoid potential collisions. Here, features of driving behavior and methods for assessing the risk of collision were investigated for a right turn at an intersection in left-hand traffic, which is a typical vehicle-on-pedestrian accident scenario. The results showed that pedestrian-collision risk can be evaluated from how the driver slows the vehicle and where the driver looks while turning during the maneuver. Moreover, pedestrian-collision risk could be predicted based on driving behavior upon commencement of steering when making an across-traffic turn.     

公路平面交叉口驾驶行为研究

面对目前公路平面交叉口交通事故严重的现状,为了公路平面交叉口交通事故致因分析和加深对交叉口驾驶行为的了解程度,探讨了公路平面交叉口驾驶行为理论,总结了目前道路交通系统交通事故原因及其导致事故所占比例,分析了交叉口交通系统,给出了影响交叉口驾驶行为的具体因素。在驾驶行为三阶段基础上,研究了交叉口驾驶行为模型,分析了交叉口驾驶行为特征和交叉口驾驶行为与交通事故发生的一般关系。并针对驾驶失误和不安全驾驶行为给出了防止措施,以减少交叉口交通事故。     

基于事故鉴定的信号控制路口电动自行车事故特征分析

电动自行车已成为群众出行的重要交通工具。随着电动自行车保有量的快速增长,涉及电动自行车的交通事故频发。为了研究电动自行车事故特征,特别是电动自行车违反交通信号灯控制的情况,基于涉及电动自行车的路口信号灯状态鉴定数据和事故数据,对事故的时间空间分布、碰撞车辆类型、骑车人个体特征、事故原因、电动自行车进入路口时信号灯状态等要素开展分析,获取了信号控制交叉口电动自行车事故特征规律,为开展此类事故预防提供参考。     

智能汽车对无信号交叉口行人的避撞控制

针对智能汽车在无信号交叉口对横穿行人的避撞问题,研究了主动转向避撞控制策略。基于多层模型预测控制方法,采用分层控制策略设计局部规划层控制器与全局跟踪层控制器,在此基础上根据交叉口处汽车与行人的轨迹特征计算人车碰撞剩余时间,改进传统人工势场法构造避撞函数,规划出既能规避交叉口内存在碰撞风险的行人又能使偏差最小的局部避撞路径,并使智能汽车在满足多项动力学约束时准确跟踪参考路径,通过搭建CarSim/Simulink联合仿真平台,结合广东省2006—2018年交通事故数据库选取对交叉口人车碰撞有显著影响的因素,设计仿真场景进行仿真分析。结果表明:智能汽车能在多个初始点完成对参考路径的跟踪,控制器对不同速度和附着条件有较高的鲁棒性,高速低附着场景中,智能汽车横向加速度小于0.4 g、质心侧偏角小于2°、前轮侧偏角小于2.5°,各约束量满足舒适性和平稳性条件;4个典型交叉口场景中,智能汽车以不同速度直行或转弯通过交叉口,均能识别横穿行人中存在碰撞风险的行人实现主动转向避撞。      

交通事故导致的高速公路拥堵状态判别方法

为量化分析不同交通事故条件下的高速公路路段拥堵情况,研究路段偶发性拥堵规律,本文构建了1个基于行程时间可靠性指标的高速公路路段拥堵判别方法。建立基于美国《公路通行能力手册》中行程时间可靠性分析方法的路段行程时间可靠性模型,并采用西南某高速公路路段实际数据校准模型。利用蒙特卡洛模拟方法生成交通事故场景,将交通事故解构为交通事故发生位置、交通事故严重程度、交通事故持续时间、交通事故发生频率4个特征,并以行程时间指数为路段拥堵量化指标,研究不同交通事故特征水平下的高速公路路段拥堵规律,并判别路段拥堵程度。研究结果表明:美国《公路通行能力手册》的行程时间可靠性分析方法具有可移植性,校准后可应用于国内高速公路路段;交通量接近饱和时,交通事故发生在出口匝道段的拥堵程度高于基本路段与入口匝道段,单车道关闭场景下的交通事故影响远高于路肩关闭场景下的交通事故;交通量接近自由流状态时,拥堵程度对严重程度不敏感;任何交通量水平下,单车道关闭场景下的交通事故持续时间一旦超过15 min,路段拥堵程度极有可能剧增。本文构建的路段拥堵判别方法,可以在精细化探究偶发性交通事故拥堵规律的同时划分路段拥堵等级,为相关部门的事故管理提供理论支撑。      

公路网事故多发段(点)判别方法研究

针对在事故多发段判别时,平面交叉节点通处理为路段,平面交叉节点的安全重要性被弱化,判别精度受到了影响的问题;并针对路网事故多发段判别时,先判别事故严重的路,再判别其上事故多发段,忽略了整体事故不严重的路上个别事故多发的段点的问题。对路网中各种等级公路进行归一化处理;根据平面交叉节点行车安全特性,把公路划分为平面交叉节点路段和普通路段;平面交叉节点路段安全影响范围的确定;从而进行路网事故多发段判别。既提高了已有动态步长过滤法等对一条公路进行事故多发段判别的精度,又解决了不同等级、不同特征公路组成的路网在同一水平下的事故多发段判别;为路网安全管理提供了基础。     

Driver behaviour modelling of vehicles at signalized intersection with heterogeneous traffic

Roads network is composed of mid-blocks and intersections. The part where two roads cross is called an intersection whereas the straight sections without intersection or any other interuptions is called mid-block. It can be observed that the vehicles on the mid-blocks tend to achieve their free-flow speeds while those at the intersections are forced to decelerate. Modelling of these sections needs to separate the intersections from mid-block. Further, drivers behave differently at these two locations. Present study attempts to separate the intersection zone of influence (IZOI) and mid-block using the manoeuvring characteristics of drivers in terms of acceleration/deceleration. These were captured through a global positioning system (GPS) device in the vehicle after sighting a red signal at the intersection. Further, this study also tried to observe whether different classes of drivers such as aggressive, normal or timid drivers, based on acceleration/deceleration behaviour exists. A junction with 1-km straight stretch in R. K. Puram New Delhi (India) was chosen for the study to find the IZOI. After identifying IZOI a video data was collected in Mumbai (India) for a stretch more than 200-m long near intersection where the red signal was visible; This enabled observing the driver behaviour more closely. Around 900 drivers of different modes were analysed to understand their behaviour. It was found that cars start reducing its speed at 160 m, motorized three-wheelers at 124 m and buses start reducing their speeds at 98 m distance from the intersection. The driver behaviours were distinct in each of the mode (Bus, Car and motorized-there-wheelers), but it emerges that the drivers cannot be classified into finite number of clusters based on the fitted normal distribution. Thus it can be seen that there are no clearly demcarcated driver behaviours irrespective of the vehicle type, such as aggressive, normal and timid categories as the intersection approaches. A normal distribution model can classify the drivers satisfactorily.     

公路三支无信号交叉口安全服务水平量化分级及应用

以公路三支无信号交叉口为研究对象、以预测事故数和实际事故数的经验贝尔估计期望值为评价指标、以交通事故预测模型为基础、以均值度量方程为标准,对公路三支无信号交叉口的安全服务水平进行了探索性分级研究。为我国公路交叉口安全分析与评价、预防与治理的实践提供强有力的理论指导与新的方法支持。     

Intersection traffic energy management system (ITEMS): formulation and validation

Urban transportation and traffic face significant challenges as cities continue to grow. Traffic intersections present unique challenges for city planners. Several studies have proposed diverse approaches to managing intersection traffic. However, these typically fail to reduce both energy and trip time for all vehicles involved. This paper proposes the intersection traffic energy management system (ITEMS). By using speed regulation of two vehicles approaching an intersection on crossroads, ITEMS is able to resolve intersection conflicts. This obviates the need for either vehicle to stop. Two-vehicle and four-vehicle scenarios were simulated. It was found that ITEMS reduced the total energy about 22% and the trip time across the intersection by almost 16% for every scenario compared to a traditional fixed-timed traffic light intersection. ITEMS was also found to be superior than always giving a green light to urban transit vehicles. Furthermore, with ITEMS, the energy and trip time of each vehicle in the scenario was reduced when compared to other strategies.     

An analysis of human factors in traffic accidents using the variation tree method

The purpose of our research is to develop a new method for the analysis of traffic accidents and to examine the effectiveness of that method. We modified the variation tree method, which was proposed in cognitive science, so that it could be applied to the analysis of the human factors in traffic accidents. We analyzed in particular accidents that occurred at intersections. The resulting diagram makes the sequence of the accident more visually clear. The diagram also illustrates the relationship between direct and indirect contributing factors.     

基于典型非机动车事故的自动驾驶汽车测试场景构建方法研究

基于交通事故卷宗、交通事故视频信息数据,研究机非混行交通环境下典型交通事故形态,构建了面向机非混行交通环境下的自动驾驶汽车测试场景,旨在针对我国较为特殊的机非混行环境下的自动驾驶汽车的测试场景及测试评价方法提供参考。本文首先分析了自动驾驶测试场景的构建需求,建立交通事故数据筛选标准,得到133例可用于构建自动驾驶汽车测试场景的机动车与非机动车交通事故数据集;其次基于《中华人民共和国道路交通安全法》行驶要求,对133例交通事故的发生地点、车辆行为、道路类型、环境光线等方面进行解构分析;最后通过聚类分析,建立了5类典型的自动驾驶测试场景模型,并分析了不同场景模型的关键要素,为实际道路测试提供理论指导。     

视线遮挡条件下面向弱势道路使用者的避撞策略研究

由于视线障碍物造成的“鬼探头”事故已经成为当前城市道路交通事故的主要类型之一。针对汽车碰撞视线遮挡条件下横穿的弱势道路使用者(VRU)的场景, 设计了1种基于碰撞时间比和安全制动距离的避撞策略, 建立车辆与VRU的交通状态数学模型, 分析“鬼探头”场景下的制动避撞临界距离。结合临界距离和车辆与VRU的碰撞时间比, 将可以避免碰撞的场景分为3种工况, 分别采用不同的制动减速度, 建立自动紧急制动避撞策略。通过Euro NCAP CPNC测试场景对该策略与传统TTC制动算法进行比较分析。结果表明, 在Euro NCAP CPNC测试场景中, 自车利用该避撞策略在理想情况下能够在更高的车速情况下完成避撞; 在不能避免碰撞的高速行驶工况中较传统TTC算法能够更加有效降低碰撞速度, 同时降低事故重伤风险和死亡风险, 提高车辆的安全性。      

节能导向的信号交叉口生态驾驶策略研究

为解决信号交叉口区域车辆频繁启停及怠速停车造成的燃油浪费和污染物排放问题,研究了以节能为导向的信号交叉口的生态驾驶策略问题.借助交叉口区域的V2I通信系统获得车辆自身定位和运动状态数据,以及信号灯状态与配时信息,对车辆所处车速引导场景及可通行性进行判定.对各场景下车辆时空运动轨迹进行分析,综合考虑交叉口上下游的燃油消耗,以平均每公里油耗最小建立统一的优化目标函数,求得生态驾驶轨迹最优解,以向驾驶员提供车速建议.利用M atlab开展的随机仿真实验表明,与不采用生态驾驶车速引导相比,采用生态驾驶车速引导至少可降低10% 以上的燃油消耗.对于划分的6种车速引导场景而言,最优生态驾驶策略在场景2下的节油效果最为显著,可达到30% ~60%;其次是场景4,可达到25% ~50%;在场景3和场景5中表现略差.从节约能源的角度而言,车辆在通过信号交叉口时应尽量避免停车等待,防止发动机长时间空转造成的燃油浪费,必要时可采取适当加减速的方式通过交叉口.      

Optimal motion control for collision avoidance at Left Turn Across Path/Opposite Direction intersection scenarios using electric propulsion

Collision avoidance at intersections involving a host vehicle turning left across the path of an oncoming vehicle (Left Turn Across Path/Opposite Direction) have been studied in the past, but mostly using simplified interventions and rarely considering the possibility of crossing the intersection ahead of a bullet vehicle. Such a scenario where the driver preference is to avoid a collision by crossing the intersection ahead of a bullet vehicle is considered in this work. The optimal vehicle motion for collision avoidance in this scenario is determined analytically using a particle model within an optimal control framework. The optimal manoeuvres are then verified through numerical optimisations using a two-track vehicle model, where it was seen that the wheel forces followed the analytical global force angle result independently of the other wheels. A Modified Hamiltonian Algorithm controller for collision avoidance that uses the analytical optimal control solution is then implemented and tested in CarMaker simulations using a validated Volvo XC90 vehicle model. Simulation results showed that collision risk can be significantly reduced in this scenario using the proposed controller, and that more benefit can be expected in scenarios that require larger speed changes.     

基于二元Logistic模型的城市道路交通事故严重程度分析

为从多维度精准剖析影响城市道路交通事故严重程度的因素，选取了我国某城市2018—2020年交通事故数据库中的4 587条数据作为研究对象，基于二元Logistic模型，从人、车、路、环境这4个方面，分别针对财产损失事故、伤人事故、死亡事故建立了模型。深入分析了道路物理隔离位置、路侧防护设施类型等因素对事故严重程度的影响，并利用Hosmer-Lemeshow检验和一致性检验对模型有效性进行验证。结果表明:①道路物理隔离的空间位置对事故严重程度有显著影响，仅布设中心隔离设施发生死亡事故的概率是同时布设中央和机非隔离的2.304倍。在有中心隔离设施的高等级道路中，增设机非隔离设施能有效降低事故发生的概率。②路侧防护设施类型为行道树、绿化带时，发生死亡事故的概率分别是金属护栏的1.982倍、1.648倍。与金属护栏相比，行道树更容易引发严重事故。③夜间无路灯照明发生死亡事故的概率是夜间有路灯照明的1.808倍，夜间无路灯照明是导致死亡事故的重要因素之一。④受过高等教育的驾驶人发生财产损失事故和伤人事故的概率较高，受过中等教育的驾驶员发生死亡事故的概率较高；受过中等教育驾驶员发生死亡事故的概率是高等教育驾驶员的2.049倍。研究深入分析了影响城市道路交通事故的显著因素及其对事故的影响，为事故严重程度的精细化分析提供了理论支持，为交通规划与管理部门提供了决策依据。      

人机混驾环境下无信号交叉口自动驾驶汽车左转运动规划研究

为了使自动驾驶汽车在人机混驾环境下能安全、高效地左转通过无信号交叉口，在借鉴人类驾驶人左转时会对周围车辆驾驶意图进行提前预判的基础上，提出了一种基于周围车辆驾驶意图预测的自动驾驶汽车左转运动规划模型。首先将无信号交叉口处周围车辆的驾驶意图分为左转、右转、直行3种类型，利用相关向量机预测周围车辆驾驶意图，以概率形式输出意图预测结果并实时更新，进一步界定自动驾驶汽车与周围车辆的潜在冲突区域并判断是否存在时空冲突；接着，在充分考虑他车速度、航向及车辆到达冲突区域边界距离的基础上建立基于部分可观测马尔可夫决策过程的自动驾驶汽车左转运动规划模型，生成一系列期望加速度；最后，基于Prescan-Simulink联合仿真平台搭建无信号交叉口仿真场景，对所提左转运动规划方法进行仿真验证，将基于博弈论的运动规划方法、基于人工势场理论的运动规划方法与所提出的方法进行比较，并选取行进比例达到1所用的时间和碰撞次数作为评价指标。研究结果表明：基于相关向量机的驾驶意图预测方法可在自动驾驶汽车到达交叉口之前准确预测出他车驾驶意图；基于部分可观测马尔可夫决策过程的左转运动规划方法能够通过速度调整策略实现人机混驾环境下自动驾驶汽车与周围车辆在无信号交叉口处的交互；不同算法对比效果表明，所提左转运动规划方法在自动驾驶汽车与不同数量周围车辆交互的仿真场景下均可有效避免碰撞事故发生并提高自动驾驶汽车左转通过无信号交叉口的效率。     

Detection of vehicles on the other crossing path at an intersection: visual search performance of elderly drivers

At intersections with good visibility, such as those located in rice fields, relatively many fatal accidents involving elderly drivers happen in Japan. In the present study, the causative factor of these accidents was examined from the viewpoint of visual search performance. It was shown that a vehicle that is not on collision course can be easily detected by peripheral vision. No age related deterioration of the detection performance in the peripheral visual field was found. However, the detection of a vehicle on collision course in the peripheral visual field was more difficult for elderly drivers. No age difference in detection performance was found when drivers were allowed to actively search the road scene by central vision. It can be concluded that active checking of the periphery at intersections with good visibility is more important for elderly drivers than for young drivers.     

基于贝叶斯网络的城市平面交叉口交通事故分析

应用贝叶斯网络对城市平面交叉口交通事故进行了分析。以3 584起交通事故数据为分析依据,基于专家知识和数据融合方法建立了城市平面交叉口交通事故分析的贝叶斯网络结构,利用服从Drichlet分布的贝叶斯方法对贝叶斯网络进行了参数学习。结合网络模型,应用联合树引擎算法推断了在车辆类型、交叉口类型、交叉口控制方式和交通参与者等因素的影响下平面交叉口交通事故类型的变化。研究结果表明,在城市平面交叉口中,由自行车导致的正面碰撞事故的概率最大,为22.83%,由于交通参与者转向不当引起的侧面碰撞的概率为23.44%,同时也易导致刮擦事故的发生;交通参与者的感知判断失误导致尾随碰撞事故的概率为23.62%。