| 自动驾驶汽车基础测试场景构建研究（双语出版） |
| |
| 引用本文： | 舒红,袁康,修海林,夏芹,何杉.自动驾驶汽车基础测试场景构建研究（双语出版）[J].中国公路学报,2019,32(11):245-254. |
| |
| 作者姓名： | 舒红 袁康 修海林 夏芹 何杉 |
| |
| 作者单位： | 1. 重庆大学 汽车工程学院, 重庆 400044;2. 中国汽车工程研究院股份有限公司汽车工程研发中心, 重庆 401122 |
| |
| 基金项目： | 智能制造专项项目（2016ZXFB06002） |
| |
| 摘 要： | 针对L2/L3级自动驾驶汽车的仿真测试和封闭场地测试认证需求，结合现有L2/L3级自动驾驶汽车量产车型的主要功能特点，提出自动驾驶汽车基础测试场景群的构建方法。首先针对指定的道路交通环境，分析主车和周围交通参与者可能的相对位置和运动方向的组合，确定复杂场景群。其次分别以主车功能所确定的各个可能运动方向，依此与各干扰车辆的可能运动方向（包括任一干扰车辆不存在的情形）进行组合，组合时采用PICT组合测试工具，并添加必要的运动约束条件，选择参数组合覆盖标准自动生成全部的组合场景群。最后结合场景筛选规则，筛选出具有测试价值的覆盖各个层级及功能的基础测试场景群。采用场景构建方法，对于主车处于三车道中间车道的路段场景和无红绿灯的十字路口场景，分别构建62种和33种基础测试场景。根据驾驶人行为特性、交通规则、汽车在城市、郊区和高速公路工况下的典型车速、加减速度、横向加速度、交通事故和自然驾驶数据库的有关场景数据等，设计主车换道工况的测试用例。采用模型预测控制框架建立主车局部路径规划和控制仿真模型，并对主车危险换道场景进行仿真。研究结果表明：主车在邻车道前车大减速的情况下实现了减速换道并避免了与本车道前车和邻车道前后车的碰撞，同时跟踪到期望跟车间距，验证了该换道测试用例的有效性。
|
| 关 键 词： | 汽车工程 场景构建 组合推理 自动驾驶汽车 测试用例 |
| 收稿时间： | 2019-04-15 |
Construction of Basic Test Scenarios of Automated Vehicles（in English） |
| |
| SHU Hong,YUAN Kang,XIU Hai-lin,XIA Qin,HE Shan.Construction of Basic Test Scenarios of Automated Vehicles（in English）[J].China Journal of Highway and Transport,2019,32(11):245-254. |
| |
| Authors: | SHU Hong YUAN Kang XIU Hai-lin XIA Qin HE Shan |
| |
| Affiliation: | 1. School of Automotive Engineering, Chongqing University, Chongqing 400044, China;2. Automotive Engineering Research Center, China Automotive Engineering Research Institute Co., Ltd., Chongqing 401122, China |
| |
| Abstract: | A construction method of the basic test scenario group of automated vehicles is proposed, aimed at the test certification requirements of L2/L3 automated vehicles in simulation tests and a closed field combined with the main functional characteristics of the existing L2/L3 automated vehicle production models. Firstly, regarding specified road traffic environment, the combinations of the possible relative position and moving direction of the ego-vehicle and the surrounding traffic participants were analyzed to determine the complex scenario group. Secondly, each possible direction of movement determined by the ego-vehicle function was respectively combined with the possible movement direction of each interfering vehicle (including any situation in which the interfering vehicle does not exist) by the Pairwise Independent Combinatorial Testing (PICT) tool, necessary motion constraints were added, and the parameter combination coverage criteria were selected to automatically generate all the combined scenario groups. Finally, combined with the scenario screening rules, the basic test scenario groups covering the various levels and functions with test value were selected. Using the scenario construction method, 62 and 33 basic test scenarios kinds were constructed for the road section scene of an ego-vehicle in the middle lane of a three-lane road and an intersection scene without traffic lights. According to the driver's behavior characteristics, traffic rules, typical vehicle speed, acceleration, deceleration and lateral acceleration in urban, suburban and highway conditions, traffic accidents, and relevant scenario data of the natural driving database, test cases for ego-vehicle lane changing conditions were designed. The model predictive control framework was used to establish the ego-vehicle local path planning and control simulation model. The simulation was carried out for a dangerous vehicle lane change scenario. The simulation results show that the ego-vehicle can realize a deceleration lane change in the case of large deceleration of a front vehicle in the adjacent lane. In addition, collisions with the front vehicle in the same lane and the front and rear vehicle in the adjacent lanes can also be thus avoided, the desired following distance tracked, verifying the effectiveness of the lane change test case. |
| |
| Keywords: | automotive engineering scenario construction combined reasoning automated vehicles test case |
|
| 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 |
|
