| 减速频次催生换道意图下自动驾驶车辆自适应换道模型 |
| |
| 引用本文： | 杨敏,王立超,王建.减速频次催生换道意图下自动驾驶车辆自适应换道模型[J].中国公路学报,2022,35(11):204-217. |
| |
| 作者姓名： | 杨敏 王立超 王建 |
| |
| 作者单位： | 东南大学交通学院, 江苏南京 211189 |
| |
| 基金项目： | 国家自然科学基金项目（52072066）；江苏省杰出青年科学基金项目（BK20200014） |
| |
| 摘 要： | 科学、合理、拟人化的换道控制是实现自动驾驶车辆安全高效行驶的重要保障，已有研究主要考虑相邻车道速度差、换道间隙等要素对车辆换道控制的影响，并未考虑车辆频繁加减速导致乘车体验差而催生换道意图这一重要现象。针对该问题，设计以抗干扰能力为基础的自动驾驶车辆自适应换道调控方法，其调控过程主要包括：采用智能驾驶人模型控制自动驾驶车辆纵向驾驶行为，以减速频次为指标度量自动驾驶车辆的抗干扰能力，并将抗干扰能力引入到自动驾驶车辆换道决策过程中，模拟自动驾驶车辆因频繁加减速导致乘车体验差而产生换道意图的现象，在此基础上，提出车辆换道控制模型。然后，以智慧高速为背景，利用Netlogo构建多种自动驾驶车辆运行场景，测试所构建的自适应换道调控方法。研究结果表明：智能驾驶人模型的选用能够合理体现自动驾驶车辆换道行为对交通流的运行影响；相比于低密度车流（≤30 veh），在中高密度车流情况下（≥40 veh），自动驾驶车辆维持原有车道运行的能力较弱、换道频率较高，且过高80次·（5 min）-1]或过低10次·（5 min）-1]的抗干扰能力临界值会导致自动驾驶车辆运行速度降低至10 km·h-1，因此可以根据不同车流密度条件对自动驾驶车辆的最大抗干扰能力进行设置和调整，从而保证自动驾驶车辆的运行效率，这也从侧面证明了所提自适应换道调控方法的科学性与合理性。研究结果对于提高自动驾驶车辆换道控制的合理自主性具有重要意义，该结果进一步完善了自动驾驶车辆换道模型库，能够为自动驾驶自适应换道调控提供理论和技术支撑。
|
| 关 键 词： | 交通工程 换道调控 自适应 乘车体验 换道意图 |
| 收稿时间： | 2021-08-27 |
Adaptive Lane-changing Model for Autonomous Vehicles Under Deceleration Frequency to Induce Lane-changing Intentions |
| |
| YANG Min,WANG Li-chao,WANG Jian.Adaptive Lane-changing Model for Autonomous Vehicles Under Deceleration Frequency to Induce Lane-changing Intentions[J].China Journal of Highway and Transport,2022,35(11):204-217. |
| |
| Authors: | YANG Min WANG Li-chao WANG Jian |
| |
| Affiliation: | School of Transportation, Southeast University, Nanjing 211189, Jiangsu, China |
| |
| Abstract: | Scientific, reasonable, and anthropomorphic lane-changing control is important for safe and efficient maneuver of autonomous vehicles. Existing studies mainly consider the influence of factors such as the speed difference between adjacent lanes and lane-changing gaps on vehicle lane-changing control, but do not consider the important phenomenon of frequent acceleration and deceleration leading to poor riding experiences. To address this problem, this study designed an adaptive lane-changing control method for autonomous vehicles based on its anti-interference ability. The adaptive regulation method mainly includes an intelligent driver model (IDM) to control the longitudinal driving behavior of autonomous vehicles. The deceleration frequency was used as an indicator to measure the anti-interference ability of autonomous vehicles. Anti-interference ability is introduced into the lane-changing decision-making process to simulate the phenomenon of lane-changing intention caused by poor riding experience, and a vehicle lane-changing control model is proposed based on this concept. Subsequently, with the background of Smart Expressway, Netlogo was used to construct a variety of autonomous vehicle operation scenarios to test the adaptive lane-changing control method. The experimental results show that the selection of the IDM model can reasonably reflect the influence of autonomous vehicle lane changes on the traffic flow. Compared with low-density traffic flow (≤ 30 vehicles), the ability of autonomous vehicles to maintain the original lane is weak in the case of medium- and high-density traffic flow (≥ 40 vehicles) scenarios. The lane-changing frequency of autonomous vehicles is the highest under medium density conditions, and if the anti-interference ability is too high80 times·(5 min)-1] or too low10 times·(5 min)-1], the speed of autonomous vehicles will be reduced. Therefore, the maximum anti-interference ability can be set and adjusted under different traffic density conditions to ensure operational efficiency. This also proves the scientific rationality of the proposed adaptive lane-changing control method. This research is of great significance for improving autonomous vehicle lane-changing control and further improving the autonomous vehicle lane-changing model library, which can provide theoretical and technical support for adaptive lane-changing control of autonomous vehicles. |
| |
| Keywords: | traffic engineering lane-changing regulation and control adaptive riding experience lane-changing intention |
|
| 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 |
|
