Safe Efficient Policy Optimization Algorithm for Unsignalized Intersection Navigation

Unsignalized intersections pose a challenge for autonomous vehicles that must decide how to navigate them safely and efficiently.This paper proposes a reinforcement learning(RL)method for autonomous vehicles to navigate unsignalized intersections safely and efficiently.The method uses a semantic scene representation to handle variable numbers of vehicles and a universal reward function to facilitate stable learning.A collision risk function is designed to penalize unsafe actions and guide the agent to avoid them.A scalable policy optimization algorithm is introduced to improve data efficiency and safety for vehicle learning at intersections.The algorithm employs experience replay to overcome the on-policy limitation of proximal policy optimization and incorporates the collision risk constraint into the policy optimization problem.The proposed safe RL algorithm can balance the trade-off between vehicle traffic safety and policy learning efficiency.Simulated intersection scenarios with different traffic situations are used to test the algorithm and demonstrate its high success rates and low collision rates under different traffic conditions.The algorithm shows the potential of RL for enhancing the safety and reliability of autonomous driving systems at unsignalized intersections.

Bicycle capacity of borrowed-priority merge at unsignalized intersections in China

To investigate bicyclists＇ behavior at unsignalized intersections with mixed traffic flow, a bicycle capacity model of borrowed-priority merge was developed by the addition-conflict-flow procedure. Based on the actual traffic situation, the concept of borrowed priority, in which the majorroad bicycles borrow the priority of major-road cars to enter the intersections when consecutive headway for major-steam cars is lower than the critical gap for minor-road cars, was addressed. Bicycle capacity at a typical unsignalized intersection is derived by the addition-conflict-flow procedure. The proposes model was validated by the empirical investigation. Numerical results show that bicycle capacity at an intersection is the function of major-road and minor-road car streams. Bicycle capacity increases with increasing major-road cars but decreases with increasing minorroad cars.

Capacity of unsignalized intersection with mixed distribution headway

In order to describe the time-headway distribution more precisely in urban traffic network,the mixed distribution model was introduced which has been widely used in mathematical statistics,and a capacity model of unsignalized intersections was obtained based on gap acceptance theory.The new model is suitable for absolute and limited priority controlled conditions and can be regarded as a more general form which handles simple headway distributions including lognormal distribution,negative exponential distribution and shifted negative exponential distribution.Through analyses of the main influencing factors in this model,the proportion of free flowing and the standard variance of gaps between any two continuous following vehicles are high sensitivity with the capacity when major stream volume is low.Besides,the capacity is affected deeply by the mean value of following vehicle gaps when major stream value is fixed and the proportion of free flowing is small.At last,the observed minor stream capacity is obtained by the survey date in Changchun city,and the average relative error between the theoretical capacity proposed in this paper is 13.73%,meanwhile the accuracy increases by 16.68% compared with the theoretical value when major stream obeys shifted negative exponential distribution.

Analysis of occupation time of vehicles at urban unsignalized intersections in non-lane-based mixed traffic conditions

In India, traffic flow on roads is highly mixed in nature with wide variations in the static and dynamic characteristics of vehicles. At unsignalized intersections, vehicles generally do not follow lane discipline and ignore the rules of priority. Drivers generally become more aggressive and tend to cross the uncontrolled intersections without considering the conflicting traffic. All these conditions cause a very complex traffic situation at unsignal- ized intersections which have a great impact on the capacity and performance of traffic intersections. A new method called additive conflict flow （ACF） method is suitable to determine the capacity of unsignalized inter- sections in non-lane-based mixed traffic conditions as prevailing in India. Occupation time is the key parameter for ACF method, which is defined as the time spent by a vehicle in the conflict area at the intersection. Data for this study were collected at two three-legged unsignalized intersections （one is uncontrolled and other one is semi- controlled） in Mangalore city, India using video-graphic technique during peak periods on three consecutive week days. The occupation time of vehicles at these intersections were studied and compared. The data on conflicting traffic volume and occupation time by each subject vehicle at the conflict area were extracted from the videos using image processing software. The subject vehicles were divided into three categories： two wheelers,cars, and auto-rickshaws. Mathematical relationships were developed to relate the occupation time of different cate- gories of vehicles with the conflicting flow of vehicles for various movements at both the intersections. It was found that occupation time increases with the increasing con- flicting traffic and observed to be higher at the uncontrolled intersection compared to the semicontrolled intersec- tion. The segregated turning movements and the presence of mini roundabout at the semicontrolled intersection reduces the conflicts of vehicular movements, which ulti- mately reduces the occupation time. The proposed methodology will be useful to determine the occupation time for various movements at unsignalized intersections. The models developed in the study can be used by practitioners and traffic engineers to estimate the capacity of unsignalized intersections in non-lane-based discipline and mixed traffic conditions.

Optimization of Quantum Cost for Low Energy Reversible Signed/Unsigned Multiplier Using Urdhva-Tiryakbhyam Sutra

One of the elementary operations in computing systems is multiplication.Therefore,high-speed and low-power multipliers design is mandatory for efficient computing systems.In designing low-energy dissipation circuits,reversible logic is more efficient than irreversible logic circuits but at the cost of higher complexity.This paper introduces an efficient signed/unsigned 4×4 reversible Vedic multiplier with minimum quantum cost.The Vedic multiplier is considered fast as it generates all partial product and their sum in one step.This paper proposes two reversible Vedic multipliers with optimized quantum cost and garbage output.First,the unsigned Vedic multiplier is designed based on the Urdhava Tiryakbhyam(UT)Sutra.This multiplier consists of bitwise multiplication and adder compressors.Compared with Vedic multipliers in the literature,the proposed design has a quantum cost of 111 with a reduction of 94%compared to the previous design.It has a garbage output of 30 with optimization of the best-compared design.Second,the proposed unsigned multiplier is expanded to allow the multiplication of signed numbers as well as unsigned numbers.Two signed Vedic multipliers are presented with the aim of obtaining more optimization in performance parameters.DesignI has separate binary two’s complement(B2C)and MUX circuits,while DesignII combines binary two’s complement and MUX circuits in one circuit.DesignI shows the lowest quantum cost,231,regarding state-ofthe-art.DesignII has a quantum cost of 199,reducing to 86.14%of DesignI.The functionality of the proposed multiplier is simulated and verified using XILINX ISE 14.2.

无信控交叉口网联车辆动态碰撞风险检测与预警策略

针对无信控交叉口碰撞事故多发的问题,面向网联人工驾驶环境,研究提出了一种基于车车通信的无信控交叉口网联车辆碰撞动态风险检测与预警策略。在构建典型无信控交叉口场景的基础上,设计了一种基于多项式拟合的车辆轨迹模型;融合利用碰撞到达时间和风险暴露时间作为风险检测指标,构建了一种圆形-双圆车辆模型实现车辆碰撞风险检测;综合考虑了驾驶人异质性及其交互行为,提出一种基于博弈论和遗传算法的两级碰撞预警策略;基于SUMO搭建仿真环境对预警策略的有效性及适用性进行测试分析。结果表明,所提策略能够准确识别出所有的碰撞事件并触发预警,预警成功率达到100%;在不同比例驾驶人组成测试工况下,所提预警策略均能显著降低碰撞率和平均碰撞动能。

无信控交叉口驾驶人潜在危险感知能力评估

为量化无信控交叉口驾驶人潜在危险的感知能力,组织驾驶人实车驾驶试验,实时采集驾驶人眼动特征参数。解析熟练和非熟练驾驶人直行通过无信控交叉口模式下的扫视和注视规律,引入马尔科夫链模型,阐释驾驶人注视转移特性,揭示驾驶人视觉搜索与潜在危险感知特性的内在联系。结合熟练和非熟练驾驶人各映射指标下的统计学分析,萃取驾驶人潜在危险感知能力表征参数。基于灰色近优综合评价法,构建驾驶人在无信控交叉口场景下的潜在危险感知能力评估模型。研究表明:直行条件下,无信控交叉口下非熟练驾驶人的水平搜索广度、垂直搜索深度和扫视强度显著低于熟练驾驶人;相较于熟练驾驶人,非熟练驾驶人兼顾道路两侧信息的能力较弱,且注视概率的适时调配机制不够灵活;熟练驾驶人在无信控交叉口的潜在危险感知得分比非熟练驾驶人高31.2%;熟练驾驶人中,男性潜在危险感知能力明显强于女性,而性别对非熟练驾驶人感知绩效无显著影响。研究结果可丰富防御性驾驶相关理论体系,为无信控交叉口交通设施的优化和完善,驾驶人安全教育及评估等提供重要参考。

基于无符号距离场的空白空间跳跃体绘制加速

随着采集技术的不断发展,各个领域都产生了大量体数据,这些数据通常具有复杂且分布不均的特征,给体绘制的实时性带来了巨大挑战。体绘制的计算复杂性非常高,如何更有效地跳过空白空间,减少无效采样次数尤为重要。为此,提出一种高效的空间跳跃体绘制加速方法,根据体数据的空间分布特征,利用计算着色器预计算得到一张三维无符号距离纹理图,该图显示了当前体素到最近一个非空体素的欧氏距离。在像素着色器中执行光线投射,每次采样前通过一次查询即可跳过连续多个空白空间,有效减少了大量无效采样次数。多种不同数据集的渲染对比实验表明,所提方法的渲染速度比传统光线投射体绘制方法提升了345%,比Occupancy Map方法提升了52%,比Chebyshev Distance Map提升了33%。

无信号交叉口处基于深度强化学习的智能网联车辆运动规划

为了兼顾无信号交叉口处智能网联车辆通行效率和舒适性要求,提出基于深度强化学习的车辆运动规划算法.结合时间卷积网络(TCN)和Transformer算法构建周围车辆行驶意图预测模型,通过多层卷积和自注意力机制提高车辆运动特征捕捉能力;利用双延迟深度确定性策略梯度(TD3)强化学习算法构建车辆运动规划模型,综合考虑周围车辆行驶意图、驾驶风格、交互风险以及自车舒适性等因素设计状态空间和奖励函数以增强对动态环境的理解;通过延迟策略更新和平滑目标策略提高算法稳定性,实时输出期望加速度.实验结果表明,所提运动规划算法能够根据周围车辆的行驶意图实时感知潜在的交互风险,生成的运动规划策略满足通行效率、安全性和舒适性要求,且对不同风格的周围车辆和密集交互场景均有良好的适应能力,不同场景下成功率均高于92.1%.

无信号交叉口网联车辆协同碰撞预警研究进展

为明晰无信号交叉口网联车辆协同碰撞预警研究的局限性及发展方向,系统梳理了协作式交叉口碰撞预警(CICW)的研究进展。首先,分类梳理了现有交叉口行车冲突检测方法的优势与不足;其次,总结了不同预警级别、机制和模式的适用性;再次,归纳了CICW有效性及安全性评价指标,对比了仿真、实车以及虚实融合测试的技术特征;然后,分析了驾驶人不确定性与通信不可靠对CICW的影响及优化思路;最后,对CICW的发展方向进行了展望。结果表明:行车风险场模型是解决现有CICW冲突检测方法中综合交通风险、冲突严重程度表征问题的可行方案,但仍需进一步研究适当的环境参数、风险指标及阈值的设定;CICW应用场景下的驾驶人驾驶特征的建模、预测和在线识别,以及CICW预警模式的自适应构建是设计可靠有效CICW应用的可行解决思路;实现对CICW的全方面客观评估,有赖于研究搭建综合评价机制及大规模试验平台;车联网通信不可靠严重影响CICW的有效性,需要进一步研究信道拥塞控制机制与基于通信故障/失效预测的CICW容错机制。

车路协同下的无信号交叉口车辆通行调度策略

无信号交叉口车辆通行调度问题是智能交通领域的研究重点,由于车辆通行顺序决策问题的解空间随着车辆数增加而指数级增长,在保证实时性的同时找到较优通行顺序成为无信号交叉口通行调度的一大问题。针对该问题提出一种基于自适应蒙特卡罗树搜索算法的无信号交叉口车辆通行调度方法,采用分层式框架,上层集中式顺序决策,下层分布式轨迹规划。首先,建立基于冲突点的交叉口模型,将网联车加入到待搜索队列中,根据交叉口中的车辆通行特点设计通行顺序的蒙特卡罗树搜索流程,以总通行时间为指标建立评价函数,然后针对不同交通情景设计自适应探索系数及其他超参数,使算法在求解不同车辆数时以及搜索的不同时期保持最佳状态。轨迹规划环节以加速度二范数为目标函数,以速度、加速度以及始终点位置等为约束,建立最优控制命题求解车辆轨迹。最后,进行实验,结果表明该算法相较于其他算法在数值仿真和微缩平台实验中最大优化幅度分别达到33.42%和38.04%,为无信号交叉口车辆通行调度提供了一个有效解决方案。

无信号灯交叉路口智能车辆调度算法研究

针对车辆在通过无信号灯交叉路口时存在通行效率低、发生冲突碰撞概率大、通行等待时间长等问题,提出了一种优化调度算法。将交叉路口网格化。考虑车辆在每个网格的方向权值、预警权值、转向权值并进行实时动态赋权,根据权值为车辆实时更新最优路径。基于此,提出了一种基于动态权值路径的算法以寻找车辆最佳行驶路径,可以根据交叉路口车辆的运行环境灵活调整每辆车的运行路线。通过将交叉路口网格化和使用动态权值规划路径相结合,该研究为避免发生碰撞、提高通行效率提供了可行的思路。

Theoretical Analysis of Cooperative Driving at Idealized Unsignalized Intersections

Cooperative driving is widely viewed as a promising method to better utilize limited road resources and alleviate traffic congestion.In recent years,several cooperative driving approaches for idealized traffic scenarios(i.e.,uniform vehicle arrivals,lengths,and speeds)have been proposed.However,theoretical analyses and comparisons of these approaches are lacking.In this study,we propose a unified group-by-group zipper-style movement model to describe different approaches synthetically and evaluate their performance.We derive the maximum throughput for cooperative driving plans of idealized unsignalized intersections and discuss how to minimize the delay of vehicles.The obtained conclusions shed light on future cooperative driving studies.

公路无信号交叉口行人过街主动安全预警技术研究应用

农村公路的本质决定了其接入干线公路存在“点多、面广、分布散”的特征。而现状交叉口普遍存在视认性差、视距不足、缺乏接入管理和交通管制、安全设施不完善、交通参与者安全意识相对薄弱、交通行为的盲目性和随意性大等问题,给道路交通安全埋下了安全隐患,一直是农村公路交通安全管理中的痛点和难点。近年来,针对农村公路交叉口警示的建设越来越多,管理的投入也越来越大,但很多都是常规的静态警示功能。传统静态防范手段无法感知判断道路交通参与者的运行信息,仅通过静态信息放大的方式往往会给驾驶员带来信息疲劳,无法达到及时提醒的目的。文章通过利用智能感知、计算机等技术,研究开发公路无信号交叉口行人过街主动安全预警技术,对交叉口的人、车进行主动探测、提前预警,提高交叉口的交通安全。

基于最大团理论的自治交叉路口控制方法

现有自治交叉路口控制策略缺乏前瞻性、容易发生死锁导致控制系统执行效率过低,针对这一问题,设计了一种基于最大团理论的车路协作自治交叉路口控制策略。首先,对车辆行驶的时空轨迹进行建模,构建可以描述车辆间行驶冲突关系的冲突矩阵;其次,将冲突矩阵转换成冲突关系图,通过设立3个求解阶段,求解冲突关系图中最大团的补集,将其作为接受车辆预约请求集合,使得每个批处理周期内更多车辆成功预约通过交叉路口的请求,在保证车辆在交叉路口行驶安全的基础上,提高了交叉路口的通行效率。仿真实验结果表明:相比于先来先服务控制策略、交通信号控制策略及基于禁忌搜索的控制策略,车辆的平均等待时间分别减少40%、17%及8%,单位时间内交叉路口的通过车辆数分别提升30%、18%及9%,证明了该策略的有效性,不仅能够提高交叉路口的吞吐量,还能有效降低车辆平均等待时间。

基于改进麻雀搜索算法的无信号交叉路口车辆调度优化

本文将无信号交叉路口内部区域离散化为多个路权点,并将车辆右转弯与行人或非机动车发生碰撞造成交通事故时所占的路权点设为“故障点”,故障点有一个至多个,本文研究无信号交叉路口在发生车辆故障时的通行效率问题.选择麻雀搜索算法提高车辆调度的通行效率,但是该算法存在前期易陷入局部最优值而后期寻优精度不高等问题,为解决此问题,引入自适应学习参数和等级反向学习的改进策略,提出基于自适应参数和等级反向学习的麻雀算法(ALSSA).选取13个基准测试函数以及Wilcoxon秩和检验P值验证ALSSA的有效性,结果表明,改进的麻雀搜索算法与其他算法相比,全局搜索能力、寻优精度等都有较大提升.最后,计算双向两车道、双向四车道、双向八车道不同车流量下的最优通行时间.

C语言中unsigned、类型转换与程序的可移植性

本文针对unsigned类型修饰符展开讨论,对unsigned系列扩展类型的存储模式、类型转换及其由此产生的程序移植性问题等进行了深入的研究。

Analyzing Pedestrian Fatality Risk in Accidents at Mid-Blocks

Objective: Pedestrian safety is considered as one of the greatest concerns, especially for developing countries. In the year of 2015, about 48% pedestrian accidents with 56% fatalities occurred at mid-blocks in Beijing. Since the high frequency and fatality risk, this study focused on pedestrian accidents taking place at mid-blocks and aimed at identifying significant factors. Methods: Based on total 10,948 crash records, a binary logit model was established to explore the impact of various factors on the probability of pedestrian’s death. Furthermore, first-degree interaction effects were introduced into the basic model. The Hosmer-Lemeshow goodness-of-fit test was used to assess the model performance. Odds ratio was calculated for categorical variables to compare significant accident conditions with the conference level. Variables within consideration in this study included weather, area type, road type, speed limit, pedestrian location, lighting condition, vehicle type, pedestrian gender and pedestrian age. Results: The calibration results of the model show that the increased fatality chances of an accident at mid-blocks are associated with normal weather, rural area, two-way divided road, crossing elsewhere in carriageway, darkness (especially for no street lighting), light vehicle, large vehicle and male pedestrian. With road speed limit increasing by 10 km/h, the probability of death accordingly increases by 46%. Older victims have higher chances of being killed in a crash. Moreover, three interaction effects are found significant: rural area and two-way divided, rural area and crossing elsewhere as well as speed limit and pedestrian age. Conclusions: This study has analyzed police accident data and identified factors significant to the death probability of pedestrians in accidents occurred at mid-blocks. Recommendations and improving measures were proposed correspondingly. Behaviors of different road users at mid-blocks should be taken into account in the future research.

车路协同无信号交叉口的矩形冲突检测及消解

针对车路协同环境下的冲突问题,建立了以系统反应时间代替驾驶员反应时间的自动驾驶车辆制动距离模型,以此作为安全距离改进了矩形冲突检测模型,并根据轨迹优化的研究思路,提出了以矩形冲突检测模型为基础的冲突消解算法,对非通行优先权车辆进行速度引导,避免车辆冲突。在车联网开源框架Veins的基础上,将交通仿真器SUMO和网络仿真器OMNeT++双向耦合,并对冲突检测模型与消解模型进行验证。仿真结果显示,该冲突检测及消解模型具有可行性,与传统无信号交叉口四路停车让行规则相比,模型中的速度引导方案能减少合流冲突车辆8.6%的平均行驶时间,减少交叉冲突车辆8.3%的平均行驶时间;合流冲突和交叉冲突中车辆的平均速度分别提高了61.4%和105.0%。在实际应用中,冲突消解模型可以为不同速度范围内的自动驾驶车辆提供速度参考,降低无信号交叉口车辆发生碰撞的概率,提高无信号交叉口的通行效率。

无控平交口实时风险预警系统对驾驶人行为的影响分析

为检验无控平交口实时风险预警系统(RTCRWS)的警示效果、了解其对驾驶人行为特性的影响;搭建3个道路交叉口虚拟仿真场景进行驾驶模拟实验,获取T_(SDNN)、T_(RMSSD)、N_(LF-norm)、N_(LF/HF)等生理行为数据,采集车速均值、车速标准差和车速变异系数等驾驶操作行为数据。统计分析了被试驾驶人行为特性,研究了RTCRWS对驾驶人行为的影响,应用因子分析方法对不同交通条件下RTCRWS的警示效果进行了综合评价。研究结果表明:RTCRWS对驾驶人群体的警示效果良好,能够提升无控平交口处的安全水平、降低驾驶人通过无控平交口的行驶车速、缓解驾驶人的紧张情绪;不同交通条件下驾驶人对RTCRWS的警示效果存在差异,驾驶人的心理紧张程度按照轻度交通流无预警、轻度交通流有预警和重度交通流有预警的顺序依次减弱。