| 基于驾驶人不满度的高速公路自动驾驶换道决策 |
| |
| 引用本文： | 陈慧,王洁新.基于驾驶人不满度的高速公路自动驾驶换道决策[J].中国公路学报,2019,32(12):1. |
| |
| 作者姓名： | 陈慧 王洁新 |
| |
| 作者单位： | 同济大学 汽车学院, 上海 201804 |
| |
| 基金项目： | 上海汽车工业科技发展基金会项目（1739）；同济大学第十二期实验教学改革项目 |
| |
| 摘 要： | 为了实现高速公路的自由换道行为决策，并满足行车安全高效性、决策结果平稳无震荡、与运动规划模块结合引导车辆行驶等要求，提出了一种基于驾驶人不满度的换道行为决策方法。首先，根据驾驶人的速度期望建立了驾驶人不满度累积模型，并基于驾驶人速度不满累积度产生换道意图。其次，依据不同车道障碍车的运动状态，设计了2种目标车道选择策略，通过预测引擎对各个待选车道进行预测和评估，选取其中行车效率较高的车道作为目标车道，同时建立换道最小安全距离模型，用以在换道全过程中判断换道的可行性。然后，将换道行为决策的结果以目标车道的形式传递给基于改进人工势场的运动规划模块，用于运动规划模块目标的选取，以引导车辆横纵向运动。最后，在CarSim/PreScan/Simulink的联合仿真平台和硬件在环平台上建立多种测试场景，验证换道行为决策算法。试验结果表明：换道行为决策算法能够依据驾驶人速度不满累积度产生稳定的换道意图，进而根据所设计的换道策略选取具有更高行车效率的目标车道，并在换道过程中持续判断换道的可行性，以应对障碍车辆突然加减速等突发状况，保证换道过程的高效性和安全性；换道行为决策算法通过目标车道的转换，引导运动规划模块调整车辆的运动，实现跟车、换道等行为。
|
| 关 键 词： | 汽车工程 换道行为决策 驾驶人不满度 自动驾驶 高速公路 运动规划 |
| 收稿时间： | 2019-01-03 |
A Decision-making Method for Lane Changes of Automated Vehicles on Freeways Based on Drivers' Dissatisfaction |
| |
| CHEN Hui,WANG Jie-xin.A Decision-making Method for Lane Changes of Automated Vehicles on Freeways Based on Drivers' Dissatisfaction[J].China Journal of Highway and Transport,2019,32(12):1. |
| |
| Authors: | CHEN Hui WANG Jie-xin |
| |
| Affiliation: | School of Automotive Studies, Tongji University, Shanghai 201804, China |
| |
| Abstract: | In this study, a decision-making method based on drivers' dissatisfaction is proposed in order to carry out lane changes on freeways and meet the requirements of safe and efficient driving, smooth decision-making results, and combined with a motion planning module to guide the vehicle's movement. First, a model of drivers' dissatisfaction was established, which was used as the basis for generating intentions of lane changes. Second, two types of efficiency strategies were built, according to the motion states of obstacles in different lanes, and a prediction engine was designed to forecast and evaluate candidate lanes. The lane with the higher driving efficiency was then selected as the target lane. Third, the minimum safety spacing model of lane changes was applied to ensure safety throughout the whole process of lane changing. The resultant target lane was then provided to the motion planning module, based on an improved artificial potential field, to identify the goals of movement. Finally, a variety of scene tests of the integration algorithms were carried out on the CarSim/PreScan/Simulink co-simulation platform as well as on the hardware-in-the-loop platform. The simulation results demonstrate that the algorithm can generate stable lane change intentions based on the accumulation of drivers' dissatisfaction, select the target lane with higher driving efficiency, guarantee the safety of lane changes throughout the whole process, and deal with unexpected situations such as the sudden acceleration or deceleration of obstacle vehicles. Simultaneously, by changing the target lane, the motion planning module can automatically adjust the vehicle's movement to enable car following and lane changing. |
| |
| Keywords: | automotive engineering lane change decision-making drivers' dissatisfaction automated driving freeway motion planning |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《中国公路学报》浏览原始摘要信息 |
|
点击此处可从《中国公路学报》下载全文 |
|
