嵌入路网图模型的自动驾驶场景描述语言

深度学习的快速发展带动着自动驾驶技术的迅速进步.深度学习感知模型在识别准确率逐步提升的同时,也存在鲁棒性和可靠性不足等隐患,需要在大量场景下进行充分测试以确保达到可接受的安全标准.基于场景的仿真测试是自动驾驶技术的核心和关键,如何描述和生成多样化仿真测试场景是需要解决的关键问题之一.场景描述语言能够描述自动驾驶场景并在虚拟环境中实例化场景获取仿真数据,但现有的场景描述语言大都缺少对于场景道路结构的高层抽象和描述.提出路网属性图来表示路网中抽象出的实体及他们的关系,并设计能简洁描述场景路网结构的语言SceneRoad. SceneRoad可以基于描述的场景道路结构特征构建路网特征查询图.这样,在路网中搜索符合描述的场景道路特征的问题被抽象为路网图上的子图匹配问题,该问题可用VF2算法求解.进一步地,将SceneRoad作为扩展集成到Scenic场景描述语言中.使用拓展后的语言随机生成大量多样的静态场景并构建仿真数据集.仿真数据集的统计信息表明生成的场景具有丰富的场景多样性.不同感知模型在真实和仿真数据集上的训练测试结果表明,模型在两个数据集上的表现呈正相关,意味着模型在仿真数据集上的评估...     

基于动态门控特征融合的轻量深度补全算法

稠密深度图在自动驾驶和机器人等领域至关重要,但是现今的深度传感器只能产生稀疏的深度测量,所以有必要对其进行补全.在所有辅助模态中, RGB图像是常用且易得的信息.现今的许多方法都采用RGB和稀疏深度信息结合进行补全.然而它们绝大部分都是利用通道拼接或逐元素求和简单的对两种模态的信息进行融合,没有考虑到不用场景下不同模态特征的置信度.提出一种以输入深度稀疏分布为指导,结合双模态信息量的动态门控融合模块,通过动态产生融合权重的方式对两个模态特征进行更高效的结合.并且根据不同模态的数据特征设计了精简的网络结构.实验结果表明所提出模块和改进的有效性,提出的网络在两个有挑战性的公开数据集KITTI depth completion和NYU depth v2上,使用了很少的参数量达到了先进的结果,取得了性能和速度的优秀平衡.     

Development of vehicle maneuvering system for autonomous driving

Recently, autonomous driving technology has received significant interest, and consequently, various autonomous driving algorithms have been developed by researchers. An experimental (conventional) vehicle verifies the strategy for autonomous driving by testing the performance and feasibility of the corresponding algorithms. Implementation of the autonomous driving strategy in the conventional vehicle system requires modification of the electronic control unit (ECU); however, this is limited by security issues and the high cost of development.In this paper, a maneuvering system that can control the steering angle, braking force, and accelerating signal so that autonomous driving does not operate via the ECU is proposed. Because the steering angle is an important factor in terms of the vehicle motion behavior, a robust angle controller based on the disturbance observer is adopted. Moreover, to ensure safety and user convenience, a user intention recognition algorithm is proposed. Based on this algorithm, the maneuvering system can be disabled during control without using any perception sensor. Finally, several experiments are conducted to evaluate the functions of the maneuvering system. A scenario-based driving test is also performed to verify the feasibility of the maneuvering system.     

碳化硅器件挑战现有封装技术

碳化硅(SiC)器件的新特性和移动应用的功率密度要求给功率器件的封装技术提出了新的挑战。现有功率器件的封装技术主要是在硅基的绝缘栅双极晶体管(Insulated Gate Bipolar Transistor,IGBT)和金属氧化物半导体场效应晶体管(Metal Oxide Semiconductor Field Effect Transistor,MOSFET)等基础上发展起来的,并一直都在演进,但这些渐进改良尚不足以充分发挥SiC器件的性能,因而封装技术需要革命性的进步。在简述现有封装技术及其演进的基础上,主要从功率模块的角度讨论了封装技术的发展方向。同时讨论了功率模块的新型叠层结构以及封装技术的离散化、高温化趋势,并对SiC器件封装技术的发展方向做出了综合评估。     

基于曲线拟合的毫米波雷达安装角度校准方法

毫米波雷达的安装角度校准是雷达正常使用并与摄像头进行数据融合的重要前提,雷达安装角度偏 转过大会导致雷达数据与摄像头数据融合失败,影响高级驾驶辅助系统(Advanced Driving Assistance System,ADAS) 的正常使用。文中提出一种基于曲线拟合的毫米波雷达安装角度校准方法,当车辆在道路上行驶时辅以车辆输入 的车速和偏航角信息,通过2000 个静止点进行曲线拟合得到雷达需要补偿的角度。相较于选择有护栏的道路进行 绕行和在标定场地部署角反射器进行安装角度校准的方法,这种方法适用的场景种类更多并且校准时间从15 分钟 以上缩短为5 分钟以下。经过实验验证,在花费更短时间完成校准后,校准精度与其它自校准方法相同为±5°。     

Efficient graph attentional network for 3D object detection from Frustum-based LiDAR point clouds

LiDAR-based 3D object detection is important for autonomous driving scene perception, but point clouds produced by LiDAR are irregular and unstructured in nature, and cannot be adopted by the conventional Convolutional Neural Networks (CNN). Recently, Graph Convolutional Networks (GCN) has been proved as an ideal way to handle non-Euclidean structure data, as well as for point cloud processing. However, GCN involves massive computation for searching adjacent nodes, and the heavy computational cost limits its applications in processing large-scale LiDAR point cloud in autonomous driving. In this work, we adopt a frustum-based point cloud-image fusion scheme to reduce the amount of LiDAR point clouds, thus making the GCN-based large-scale LiDAR point clouds feature learning feasible. On this basis, we propose an efficient graph attentional network to accomplish the goal of 3D object detection in autonomous driving, which can learn features from raw LiDAR point cloud directly without any conversions. We evaluate the model on the public KITTI benchmark dataset, the 3D detection mAP is 63.72% on KITTI Cars, Pedestrian and Cyclists, and the inference speed achieves 7.9 fps on a single GPU, which is faster than other methods of the same type.     

面向无人驾驶地图更新的安全信任管理方案

地图的实时性和准确性在无人驾驶车辆导航中发挥着重要作用。基于卫星图像的数字地图能够保证地图中道路信息的准确性,但是根据相关路况信息进行实时更新时,由于车辆用户数较多且来源复杂,因此无法确保数据的安全性和可靠性。提出一种基于区块链的认证与信任管理方案。利用无证书签名技术实现车辆的匿名认证,保证车辆身份的可验证性和不可否认性。设计基于区块链的信任管理机制,通过存储在区块链上的历史信任值计算路况消息的可信度,同时将可靠的路况信息发送给地图更新服务器,根据车辆提供信息的真实与否更新车辆信誉值,实现车辆信誉的同步与更新。从计算开销、通信开销和信任管理功能等方面验证了该方案的有效性。仿真结果表明,该方案具有较高的地图更新效率和防骗能力,能够确保数据来源的安全性和可靠性,为无人驾驶车辆提供实时路况更新服务。     

车联网信息对驾驶员视觉注意与行为的影响

目的探讨不同驾驶情景下车联网信息呈现形式及复杂度对驾驶过程和绩效的影响。方法采用3(信息复杂度:无、简单、复杂)*2(信息通道:视、听)*3(驾驶场景:纵向、横向、转角)多因素混合设计,以驾驶员视觉注意指标、驾驶行为指标为因变量,共招募60名驾驶员,分两组。结果 (1)车联网信息以听觉通道呈现会增加行车稳定性及方向盘控制稳定性,且不会损害驾驶员的视觉注意行为,视觉则相反;(2)信息简单也会增加行车稳定性及方向盘控制稳定性。结论横向与纵向路径以听觉呈现简单信息有利于促进安全驾驶行为,转角路径中应避免车联网信息。     

轻音乐对驾驶疲劳的双向影响

目的本实验探索驾驶负荷累积与轻音乐对驾驶疲劳的影响。方法采用2(驾驶负荷累积)×2(音乐条件)×3(模拟驾驶阶段)混合设计,对32名驾驶员被试进行模拟驾驶实验,采集驾驶员驾驶速度和他们的垂直搜索广度与瞳孔直径。结果第45-60 min时,对于无驾驶负荷累积的被试,轻音乐组车速显著快于无音乐组,轻音乐组垂直搜索广度显著大于无音乐组;对于有驾驶负荷累积的被试,轻音乐组的瞳孔直径小于无音乐组(边缘显著),轻音乐组的车速慢于无音乐组。结论轻音乐对无累积疲劳的驾驶员具有唤醒作用,加快他们的车速,扩大垂直搜索广度,对驾驶绩效具有积极的影响;轻音乐对有累积疲劳的驾驶员会产生催眠作用,缩小瞳孔直径,车速减慢,对驾驶绩效具有消极的影响。     

Traffic accidents involving fatigue driving and their extent of casualties

The rapid progress of motorization has increased the number of traffic-related casualties. Although fatigue driving is a major cause of traffic accidents, the public remains not rather aware of its potential harmfulness. Fatigue driving has been termed as a “silent killer.” Thus, a thorough study of traffic accidents and the risk factors associated with fatigue-related casualties is of utmost importance. In this study, we analyze traffic accident data for the period 2006–2010 in Guangdong Province, China. The study data were extracted from the traffic accident database of China's Public Security Department. A logistic regression model is used to assess the effect of driver characteristics, type of vehicles, road conditions, and environmental factors on fatigue-related traffic accident occurrence and severity. On the one hand, male drivers, trucks, driving during midnight to dawn, and morning rush hours are identified as risk factors of fatigue-related crashes but do not necessarily result in severe casualties. Driving at night without street-lights contributes to fatigue-related crashes and severe casualties. On the other hand, while factors such as less experienced drivers, unsafe vehicle status, slippery roads, driving at night with street-lights, and weekends do not have significant effect on fatigue-related crashes, yet accidents associated with these factors are likely to have severe casualties. The empirical results of the present study have important policy implications on the reduction of fatigue-related crashes as well as their severity.