自动驾驶智能系统测试研究综述

随着人工智能技术的深入发展,自动驾驶已成为人工智能技术的典型应用,近十年来得到了长足的发展,作为一类非确定性系统,自动驾驶车辆的质量和安全性得到越来越多的关注.对自动驾驶系统,特别是自动驾驶智能系统(如感知模块、决策模块、综合功能及整车)的测试技术得到了业界和学界的深入研究.调研了56篇相关领域的学术论文,分别就感知模...     

Path Planning for Autonomous Underwater Vehicles

Efficient path-planning algorithms are a crucial issue for modern autonomous underwater vehicles. Classical path-planning algorithms in artificial intelligence are not designed to deal with wide continuous environments prone to currents. We present a novel Fast Marching (FM)-based approach to address the following issues. First, we develop an algorithm we call FM* to efficiently extract a 2-D continuous path from a discrete representation of the environment. Second, we take underwater currents into account thanks to an anisotropic extension of the original FM algorithm. Third, the vehicle turning radius is introduced as a constraint on the optimal path curvature for both isotropic and anisotropic media. Finally, a multiresolution method is introduced to speed up the overall path-planning process     

AUV自主导航航位推算算法的分析研究

对AUV (Autonomous Underwater Vehicle)自主导航的航位推算算法做了进一步研究并加以改进，以提高其自主导航精度．然后，利用AUV湖试所获得的数据，对本文提出的修正算法进行了验证．结果表明， AUV的自主导航精度得到很大提高，可以用于修正原来的自主导航算法．     

汽车驾驶员模糊控制模型研究

建立合适的汽车方向控制驾驶员模型是人-车-路闭环系统最重要的环节之一.传统驾驶员模型的建立依赖于汽车系统传递函数,但汽车动力学控制系统具有强非线性,其传递函数不易确定,这使驾驶员模型难以建立.在"预瞄最优曲率模型"的基础上对驾驶员校正环节采用模糊控制,对包括"魔术公式"轮胎模型在内的汽车模型建立带有自调整因子的加速度反馈模糊控制驾驶员模型.该模型不需要知道汽车系统精确的传递函数,采用模糊逻辑推理直接模拟人的操纵过程来进行控制.仿真结果表明,所建立的模糊控制驾驶员模型很好地描述了驾驶员的方向控制行为,为人-车-路闭环系统的进一步研究和智能车辆自动驾驶控制提供了可行的途径.     

基于PXI图像采集卡的WDM驱动程序设计

以基于CPCI的高速图像数据采集卡为例,介绍了PXI总线在高速数字通信中的一种应用及在Windows2000下的Compact PCI设备WDM驱动程序运行机理及编程要点;针对微软提供的驱动程序开发工具包DDK给出了基于一定实时要求的程序实例;重点介绍了驱动程序中的资源分配、中断响应、DMA聚拢/分散传输及Win32应用程序与驱动程序通信的关键问题;实验表明该驱动程序实现了预期的要求.     

高速数据采集卡WDM驱动程序的开发

WDM是Microsoft Windows 2000/XP操作系统的标准驱动程序模型。文中在简要介绍一款基于PCI总线高速数据采集卡硬件设计的基础上,详细阐述了基于WDM模型的驱动程序设计。并结合项目的实际情况,对设计中一些策略的考虑和编程技巧做了讨论,给出了利用DDK开发WDM驱动程序的实例。     

智能车逆向超车控制算法

针对双向车道因受限于道路条件及交通特性仅能借用对向车道完成超车（逆向超车）的问题,通过采用车联网以及车载传感器获取环境车辆的速度、加速度等全局信息,将多车场景中各个实体所造成的影响纳入超车决策中,从而提出一种基于图搜索和模型预测控制（Model Predictive Control, MPC）的逆向超车控制方法。首先,根据车车通信获取的全局信息,结合非合作博弈,对各车在整个时段内的行为进行预测,并根据预测情况对道路的各个区域进行安全评估,评估依据为该区域在下一时刻出现车辆的概率。对道路完成评估后,得到碰撞概率热区图,之后采用A*算法搜索安全路径,根据安全路径完成目标车辆的轨迹规划,并设计模型预测控制器来对主车进行实时控制,使车辆按照既定轨迹行驶。最后,借助Carsim与MATLAB/Simulink搭建联合仿真平台,对提出的算法进行验证。仿真实验结果表明,该模型的控制误差最大不超过0.15 m,平均误差率约为1.7%,能实现对车辆的精准控制,保证被控车辆安全完成逆向超车。     

基于贝叶斯预测的自主车辆避障方法研究

针对在动态环境中自主车辆对于动态障碍物信息很难具有先验知识的问题,研究了动态贝叶斯网络模型对机动目标运动状态进行贝叶斯预测的推理机理,提出了一种基于贝叶斯预测进行自主车辆避障路径规划控制方法;该规划方法在VORONOI图法基础上,融合了对自主车辆和周围环境之间的位置荚系的贝叶斯预测,一旦预定任务的动态环境发生重大变化,它可以产生机动目标沿某方向前进信息的预测先验知识,通过局部多次重规划生成避障路径,直至自主车辆完成既定任务;仿真实验证明了该规划控制方法可有效帮助自主车辆在不确定环境中实施避障策略.     

自动驾驶汽车的高效对抗性场景测试方法研究

在自动驾驶安全性的研究和应用中,测试里程长、暴露危险场景单一的问题使自动驾驶安全性能的提升受到限制。使用对抗性场景进行测试被认为是解决上述问题的重要手段,然而,现有研究采用通用的优化算法作为框架,将大量计算资源浪费在对参数空间的探索过程中,效率低下。在计算成本的约束下,这些算法甚至无法在更复杂的环境中测试出足够多、足够丰富的失效样本。复杂环境中的对抗性场景测试面临三大挑战：信息匮乏;对抗性样本在庞大的参数空间中稀疏分布;搜索过程中探索与利用难以平衡。该文从这三大挑战出发,提出一种高效的对抗性场景测试框架,通过代理模型来获取更多关于参数空间的信息,精选小样本,以打破庞大空间中稀疏事件的制约,对未知区域和对抗性样本附近的目标进行有针对性的搜索和更新,以实现探索和利用的平衡。实验证明,该文提出方法的搜索效率是随机采样的4倍,与通用遗传算法相比,效率提升一倍以上,在有限的仿真测试次数下,生成了更多容易使被测自动驾驶系统失效的对抗性测试用例。特别地,该文提出的方法能够找出许多离群的对抗性样本,揭示出现有算法无法识别的失效模式。此外,该文提出的方法能够快速、全面地定位出被测算法的脆弱场景,为自动驾...     

电动汽车功率驱动单元设计

以电动汽车用永磁同步电机为控制对象,介绍了功率驱动单元的设计方法,包括基本结构、功率器件的选取及其驱动电路设计,以及保护电路的结构和参数设计;该方法选用MOSFET作为功率器件,降低了功率驱动单元设计的成本;采用的保护电路设计方法从参数设计入手,并通过电路布局的优化,极大地降低了MOSFET漏极和源极间dudt引发的干扰,提高了功率驱动单元的性能;应用结果表明,采用该方法设计的功率驱动单元性能优良,已成功应用于电动汽车驱动系统中。     

Field Testing of Moving Short‐baseline Navigation for Autonomous Underwater Vehicles using Synchronized Acoustic Messaging

This paper presents the results from field testing of a unique approach to the navigation of a fleet of autonomous underwater vehicles (AUVs) using only onboard sensors and information provided by a moving surface ship. The approach, considered moving short‐baseline (MSBL) navigation, uses two transponders mounted on a single surface ship that alternately broadcast acoustic messages containing one of the parameters of the kinematic state of the surface ship. The broadcasts are initiated according to a predefined schedule so that the one‐way travel time (OWTT) of the acoustic messages may be used to determine the range to the transponder. Each AUV in the fleet uses the surface ship state measurements and ranges provided by the acoustic messages in two extended Kalman filters (EKFs) for state estimation. The first EKF merges the intermittent surface ship state measurements with a kinematic model to estimate the state of the surface ship. This is necessary because the presented approach uses 13‐bit acoustic messages as opposed to the more commonly used 32‐byte messages, which allow the full state to be encoded in a single broadcast. The second EKF uses the current surface ship state estimate to properly interpret the acoustic ranges, combining them with a kinematic model to estimate the state of the AUV itself. Numerous MSBL navigation experiments were compared against a more traditional approach using a long‐baseline (LBL) array of transponders and OWTT acoustic ranging. The results of all tests were verified by independent LBL measures of position.     

Modelless Guidance for the Docking of Autonomous Vehicles

A novel line-of-sight sensing-based modelless guidance strategy is presented for the autonomous docking of robotic vehicles. The novelty of the proposed guidance strategy is twofold: 1) applicability to situations that do not allow for direct proximity measurement of the vehicle and 2) ability to generate short-range docking motion commands without a need for a global sensing-system (calibration) model. Two guidance -based motion-planning methods were developed to provide the vehicle controller with online corrective motion commands: a passive-sensing-based and an active-sensing-based scheme, respectively. The objective of both proposed guidance methods is to minimize the accumulated systematic errors of the vehicle as a result of the long-range travel, while allowing it to converge to its desired pose within random-noise limits. Both techniques were successfully tested via simulations and experiments, and are discussed herein, in terms of convergence rate and accuracy, in addition to the types of localization problems for which each method could be specifically more suitable.     

A Novel Navigation Method for Autonomous Mobile Vehicles

This paper presents a novel navigation method for Autonomous Mobile Vehicle in unknown environments. The proposed navigator consists of an Obstacle Avoider (OA), a Goal Seeker (GS), a Navigation Supervisor (NS) and an Environment Evaluator (EE). The fuzzy actions inferred by the OA and the GS are weighted by the NS using the local and global environmental information and fused through fuzzy set operation to produce a command action, from which the final crisp action is determined by defuzzification. The EE tunes the supports of the fuzzy sets for the OA and the NS; therefore, the capability of the navigation method is enhanced. Simulation shows that the navigator is able to perform successful navigation task in various unknown or partially known environments, and it has satisfactory ability in tackling moving obstacles. More importantly, it has smooth action and exceptionally good robustness to sensor noise.     

Perception and Planning Architecture for Autonomous Ground Vehicles

By combining smart sensors and traversability grids with a JAUS-based component and messaging architecture, DARPA Grand Challenge finalist team CIMAR quickly developed a robust autonomous ground vehicle (AGV), a custom-built off-road vehicle. Key components included six smart sensors for detecting environmental conditions and reporting a priori data, a smart arbiter for fusing data from multiple smart sensors, and a reactive driver for providing real-time navigation planning and obstacle avoidance     

基于神经网络的自治水下机器人广义预测控制

针对自治式水下机器人高度非线性和时变性的特点,提出了一种基于神经网络的水下机器人广义预测控制策略．利用改进型Elman网络作为多步预测模型,在对网络学习算法进行改进的基础上,实现了Elman网络的在线学习,并提出了用于求解神经广义预测控制律的灵敏度公式．进行了具有神经网络在线学习功能和不具有在线学习功能的水下机器人的速度控制实验,并就预测控制效果进行了对比分析．实验结果表明,具有自适应学习功能的水下机器人速度控制法的精度要优于不具有在线学习功能的速度控制法,且当水下机器人动态特性发生变化时具有较强的自适应能力．     

Gaussian-Process-Based Real-Time Ground Segmentation for Autonomous Land Vehicles

Ground segmentation is a key component for Autonomous Land Vehicle (ALV) navigation in an outdoor environment. This paper presents a novel algorithm for real-time segmenting three-dimensional scans of various terrains. An individual terrain scan is represented as a circular polar grid map that is divided into a number of segments. A one-dimensional Gaussian Process (GP) regression with a non-stationary covariance function is used to distinguish the ground points or obstacles in each segment. The proposed approach splits a large-scale ground segmentation problem into many simple GP regression problems with lower complexity, and can then get a real-time performance while yielding acceptable ground segmentation results. In order to verify the effectiveness of our approach, experiments have been carried out both on a public dataset and the data collected by our own ALV in different outdoor scenes. Our approach has been compared with two previous ground segmentation techniques. The results show that our approach can get a better trade-off between computational time and accuracy. Thus, it can lead to successive object classification and local path planning in real time. Our approach has been successfully applied to our ALV, which won the championship in the 2011 Chinese Future Challenge in the city of Ordos.     

Secure Mobility and the Autonomous Driver

Autonomous mobility systems developed for unmanned ground vehicles may have additional benefits by enhancing system performance and reducing demands on operators for manned ground vehicles. This effort examines the potential impact of introducing autonomous mobility to control manned vehicles while operators performed secure mobility. Eleven Soldiers participated in an experimental task requiring concurrent control of a manned and an unmanned Stryker performing a road march, scanning of the local environment for targets, and planning of a reconnaissance route for a third simulated asset. The control of the manned vehicle was varied between autonomous and manual control and several speed and accuracy variables were examined for each task. Subjective measures of operator workload, stress, and motion sickness were also examined. The results support the potential benefits of incorporating autonomous mobility into manned platforms. In speed-matched conditions, autonomous mobility was associated with decreased manned vehicle mission time, faster operator reaction times to targets, greater instances of multitasking while under motion, and lower subjective operator workload measures than with manual driving. In conclusion, autonomous mobility technologies have the potential to free up resources from the vehicle operator and allow for better operator performance on tasks other than vehicle control.     

Pipeline following by visual servoing for Autonomous Underwater Vehicles

A nonlinear image-based visual servo control approach for pipeline following of fully-actuated Autonomous Underwater Vehicles (AUV) is proposed. It makes use of the binormalized Plücker coordinates of the pipeline borders detected in the image plane as feedback information while the system dynamics are exploited in a cascade manner in the control design. Unlike conventional solutions that consider only the system kinematics, the proposed control scheme accounts for the full system dynamics in order to obtain an enlarged provable stability domain. Control robustness with respect to model uncertainties and external disturbances is reinforced using integral corrections. Robustness and efficiency of the proposed approach are illustrated via both realistic simulations and experimental results on a real AUV.