共查询到20条相似文献,搜索用时 228 毫秒
1.
《汽车工程》2021,(4)
为提升智能驾驶系统的纵向跟车性能,本文构建了一种基于深度强化学习的驾驶员跟车模型。首先,设计了跟车场景截取准则并从自然驾驶数据中筛选出符合条件的典型跟车场景,并对其数据进行统计分析,即采用相关系数法分析了车间距、相对速度和车头时距等因素对驾驶员跟车行为的影响机理,得出驾驶员跟车行驶过程的行为特性及其影响因素。接着,基于深度确定性策略梯度算法建立了驾驶员跟车模型,将驾驶员跟车轨迹数据集输入到模拟跟车环境中,让智能体从经验数据中学习驾驶员的决策行为。最后,以原始工况数据为基准,对基于深度强化学习的跟车模型进行对比仿真验证,结果表明所构建的驾驶员跟车模型具有良好的跟踪性能,能真实地复现驾驶员的跟车行为。 相似文献
2.
3.
针对网联车辆行驶过程中容易出现旁车道的人类驾驶员并线进入车队的问题,本文中提出了一种考虑旁车并线行为的跟车策略,并设计了分布式应用的分层控制系统。首先分析了所提出的跟车策略的合理性,并构建了考虑延时与误差反馈的联网巡航控制(CCC)系统;接着在频域范围内分析了不同控制增益参数对系统稳定性的影响,仿真结果验证了多车队列行驶稳定性;最后搭建测试平台进行实车试验。结果表明:旁车并线时,CCC控制系统可快速实现车辆的制动并保证队列的稳定性,所提出的跟车策略可提升车辆的乘车舒适性与交通系统的安全性。 相似文献
4.
当前不少前向碰撞预警系统以预警距离作为预警的特征量对驾驶人进行预警,因此,提高对跟车距离的预测准度能够直观有效提高该前向碰撞预警系统的预警能力。研究通过驾驶模拟器构建跟车场景,收集了41名驾驶员的跟车行为数据,按照3:1的比例将试验数据划分为训练集和测试集。将驾驶人的跟车距离与速度作为长短期记忆模型的输入,跟车距离作为模型的输出,对驾驶人的跟车距离进行了预测分析研究。结果表明,利用该数据集的模型能够很好的预测驾驶人的跟车行为,泛化性能较好,没有过度拟合现象。并且通过输入不同时间窗口长度的测试集发现,随着测试集长度的降低,预测结果的误差会更大。能够为提高前向碰撞预警系统的精准度提供理论支持,从而增加驾驶员对预警系统的接受度。 相似文献
5.
为了获取驾驶员跟车行为特性并以此为基础设计自适应巡航控制系统,建立驾驶员控制增益随车速变化的动态跟车模型。引入驾驶员追踪误差敏感度,定量分析控制增益与车速的动态变化关系。为了准确描述驾驶员行为特性,定义速度误差敏感系数SVE(Sensitivity to Velocity Error)和距离误差敏感系数SDE(Sensitivity to Distance Error),基于非线性优化算法求解模型参数。最后通过Matlab搭建自适应巡航系统,进行仿真试验,并与驾驶员试验结果对比,验证控制算法。 相似文献
6.
7.
8.
9.
10.
11.
S. Li J. Wang K. Li X. Lian H. Ukawa D. Bai 《International Journal of Automotive Technology》2010,11(1):81-87
In order to capture drivers’ car-following characteristics and apply this information to the design of an Adaptive Cruise
Control algorithm, this paper builds a driver car-following model with vehicle speed-dependent control gains. Proposed for
use with heavy-duty truck drivers, we introduced the concept of driver sensitivity to tracking errors, identified driver’s
sensitivity to tracking errors and analyzed quantitatively the relationship between control gain and vehicle speed. To model
the driver characteristics precisely and concisely, a SVE/SDE (Sensitivity to Velocity Error/Sensitivity to Distance Error)
based linear car-following model was built and a nonlinear optimization algorithm was adopted to identify the model parameters.
When validating the model accurancy, we proposed a comparative verification method based on hypothesis-testing theory here
to reduce the influence of randomicity in the drivers’ manipulation. The modeling and verification indicate that the proposed
car-following model is superior to the tranditional linear car-following model, but its structure still approximates linear,
which implies it is applicable for the design of a vehicular following controller. 相似文献
12.
In this paper, a proposed car-following driver model taking into account some features of both the compensatory and anticipatory model representing the human pedal operation has been verified by driving simulator experiments with several real drivers. The comparison between computer simulations performed by determined model parameters with the experimental results confirm the correctness of this mathematical driver model and identified model parameters. Then the driver model is joined to a hybrid vehicle dynamics model and the moderate car following maneuver simulations with various driver parameters are conducted to investigate influences of driver parameters on vehicle dynamics response and fuel economy. Finally, major driver parameters involved in the longitudinal control of drivers are clarified. 相似文献
13.
14.
为明确跨江大桥的跟驰行为特征以及驾驶模式,在重庆菜园坝大桥展开了30位被试的小客车实车驾驶试验,使用华测航姿测量系统和前视碰撞预警系统Mobileye 630采集自然驾驶状态下汽车的连续行驶速度、车头时距和车头间距等数据。通过筛选得到了725条有效跟驰轨迹数据,对比分析发现跨江大桥与城市一般道路的跟驰行为存在一定差异性,明确了菜园坝大桥车头时距和车头间距的分布特征,并且对强跟驰(小于1.6 s)、过渡区间(1.6~2.6 s之间)以及弱跟驰(大于2.6 s)3种跟驰状态和驾驶人性别差异下的跟驰数据进行了分析。结果表明:桥梁段车头时距分布集中在1.6 s处,车头间距分布集中在18 m处;超过1/3的跟驰轨迹处于强跟驰状态,此状态下前车驾驶行为对跟驰车辆具有较强制约性;当车辆处于弱跟驰状态时,前车对于后车的约束性会随车头时距的增大而快速降低;过渡区间的设立更好地揭示了强/弱跟驰状态之间的转变并不是只有一个临界值,而是存在一个转换过程,并且其间车辆跟驰特性的变化与驾驶人本身的操作行为存在较大关联;驾驶人的性别差异对跟驰距离几乎没有影响,但男性驾驶人往往会采取更加冒险的驾驶行为,平均车头时距、车头间距以及相对速度均高于女性驾驶人。 相似文献
15.
雾环境下驾驶人行车与正常天气相比,在低能见度下视觉参照物较少,驾驶人更倾向于跟驰行驶。为研究雾环境下高速公路驾驶人跟驰行为,以真实雾环境下实车试验方式,选择多条高速公路作为试验路段,以Smart Eye眼动仪获取车辆在雾环境下高速公路驾驶人视觉参数,包含驾驶人注视区域、注视角度、注视持续时间、瞳孔直径、扫视速度以及扫视幅度等,以归一化方法对驾驶人注视重心进行分析,研究不同能见度下驾驶人的跟驰需求,并通过对雾环境下上述视觉参数进行规律总结。对雾环境下驾驶人跟驰特性进行统计及分类,将跟驰行为划分为主动、半主动、半被动以及全被动跟驰;通过分析雾区低能见度下驾驶人跟驰行驶条件,引入多维偏好理论及后悔理论,进行驾驶人跟驰决策模型构建,并基于差分法对模型进行参数标定及验证。研究结果表明:驾驶人在1次跟驰动态过程中,正常车道保持时驾驶人扫视速度较低,而当处于车道调整时,驾驶人扫视速度存在较大波动,且平均扫视速度较高,低能见度下驾驶人注视点转移速度27.0 (°)·s-1明显低于晴好天气的52.0 (°)·s-1;驾驶人在跟驰过程中,能见度对驾驶人跟驰时的视觉特征有显著影响,通过跟驰模型构建可为后续雾环境下车辆跟驰前后车距及车速预测提供理论支撑。 相似文献
16.
《Vehicle System Dynamics: International Journal of Vehicle Mechanics and Mobility》2012,50(7):1085-1102
The paper presents a curving adaptive cruise control (ACC) system that is coordinated with a direct yaw-moment control (DYC) system and gives consideration to both longitudinal car-following capability and lateral stability on curved roads. A model including vehicle longitudinal and lateral dynamics is built first, which is as discrete as the predictive model of the system controller. Then, a cost function is determined to reflect the contradictions between vehicle longitudinal and lateral dynamics. Meanwhile, some I/O constraints are formulated with a driver permissible longitudinal car-following range and the road adhesion condition. After that, desired longitudinal acceleration and desired yaw moment are obtained by a linear matrix inequality based robust constrained state feedback method. Finally, driver-in-the-loop tests on a driving simulator are conducted and the results show that the developed control system provides significant benefits in weakening the impact of DYC on ACC longitudinal car-following capability while also improving lateral stability. 相似文献
17.
车辆跟驰行为建模的回顾与展望 总被引:2,自引:0,他引:2
系统地回顾了跟驰理论60年的发展历程,依据建模思想将跟驰行为模型分为交通工程角度和统计物理角度。交通工程角度的跟驰模型包括刺激-反应类、安全距离类、心理-生理类及人工智能类模型;统计物理角度的跟驰模型包括优化速度模型、智能驾驶模型和元胞自动机模型。针对各类模型分别阐述了其建模思路、模型结构、参数标定及其扩展研究。最后,展望了跟驰行为建模的发展趋势与研究方向,为建立适合中国交通流特点的跟驰模型提供参考。 相似文献
18.
Y. Hattori K. Asano N. Iwama T. Shigematsu 《Vehicle System Dynamics: International Journal of Vehicle Mechanics and Mobility》1995,24(4):299-311
This report describes a decelerating driver-model expressed by driving mode transition in car-following situations. The assumptions for constructing the model are that decelerating strategy of a driver is classified into several simple driving modes and that a driver changs his driving modes based on his perceptible characteristics and experiential rules. Deceleration action is divided into three states; following, standing and braking, which are applied to the model. The model has two paths for driver's decelerating action, one of which is selected by the driver based on the perceptible characteristics and experiential rules. The suitability of the model has been experimentally verified. 相似文献
19.
《Vehicle System Dynamics: International Journal of Vehicle Mechanics and Mobility》2012,50(6):751-773
An errorable car-following driver model is presented in this paper. An errorable driver model is one that emulates human driver’s functions and can generate both nominal (error-free), as well as devious (with error) behaviours. This model was developed for evaluation and design of active safety systems. The car-following data used for developing and validating the model were obtained from a large-scale naturalistic driving database. The stochastic car-following behaviour was first analysed and modelled as a random process. Three error-inducing behaviours were then introduced. First, human perceptual limitation was studied and implemented. Distraction due to non-driving tasks was then identified based on the statistical analysis of the driving data. Finally, time delay of human drivers was estimated through a recursive least-square identification process. By including these three error-inducing behaviours, rear-end collisions with the lead vehicle could occur. The simulated crash rate was found to be similar but somewhat higher than that reported in traffic statistics. 相似文献
20.
为弥补传统风险评价指标对相对速度较小的跟车场景危险水平评价能力的不足,减少跟车场景中追尾事故的发生,提出了跟车场景潜在风险的概念。将假定前车以较大制动减速度减速条件下的风险称为潜在风险,并构建了代表驾驶人在潜在危险跟车场景下进行避撞操作需满足的最大反应时间的参数时间裕度(TM)。由于追尾危险工况中常见采取的避撞操作可分为制动和制动转向两大类,分别对典型制动避撞过程和制动转向避撞过程进行了分析,从而推导出分别针对2种跟车潜在危险场景的TM计算方式。通过自动筛选与人工筛选结合,获得了中国自然驾驶数据库(China-FOT)中具有中国驾驶人特点的制动避撞危险工况87个和转向制动避撞危险工况40个进行分级,并基于图像处理方法提取了前车制动开始时刻的TM值,从而得到跟车场景潜在风险两级危险域的划分。结果表明:制动避撞场景下,本车车速过低和过高时,TM值的变化均不明显;而在制动转向避撞场景中,只有速度较高时阈值才保持不变。通过对正常驾驶视频的分析,引入对比组共计正常跟车制动工况163例和正常跟车转向变道工况151例的车头时距(THW)值,其统计分析结果与支持向量机分类结果均难以清晰描述跟车场景危险水平与本车速度之间的关系。为研究跟车场景潜在风险评价在自动驾驶中的应用,对于制动避撞场景,在设定TM值不变和相对速度不变的条件下,分别对基于TM的最小相对距离和距离碰撞时间(TTC)值进行了仿真,仿真结果显示,相比于TTC而言,TM的评价稳定性受相对速度影响小,在自动驾驶跟车策略开发和避免其发生责任追尾事故中有重要应用价值。 相似文献