A Vision for Supporting Autonomous Navigation in Urban Environments

The autonomous navigation systems (ANS), such as autonomous ground vehicles (AGVs), unmanned aerial vehicles (UAVs), and unmanned submersible vehicles (USVs), and modern vehicles with actuators, sensors, and computer control perform three basic functions: context gathering using sensors, processing, and action. Most researchers have put all three functions into the ANS or the robot itself to overcome occlusions and handle the environment's dynamics. However, this causes the ANS and robotic systems to be bulky and expensive. It also impedes the introduction of vehicles with ANS in urban environments, where they must coexist with existing cars and highways. The approach presented distributes the context-gathering and processing functions using sensor networks and wireless communications technologies to reduce costs and make ANS widespread. The system uses sensors mounted on moving vehicles and stationary objects such as lampposts, traffic lights, toll plazas, and buildings to gather information at different levels     

复杂环境下多无人机协作式地面移动目标跟踪

针对多无人机(UAV)协同地面移动目标跟踪问题展开研究.提出一种基于主动感知的问题求解框架,建立多UAV协同目标跟踪问题模型;在此基础上,采用分布式无色信息滤波实现目标状态融合估计与预测;然后,基于预测目标状态,结合滚动时域控制与遗传算法设计一种多UAV在线协同航迹规划算法.仿真结果表明:结合预测目标状态在线优化UAV...     

Verification of heterogeneous multi-agent system using MCMAS

The focus of the paper is how to model autonomous behaviours of heterogeneous multi-agent systems such that it can be verified that they will always operate within predefined mission requirements and constraints. This is done by using formal methods with an abstraction of the behaviours modelling and model checking for their verification. Three case studies are presented to verify the decision-making behaviours of heterogeneous multi-agent system using a convoy mission scenario. The multi-agent system in a case study has been extended by increasing the number of agents and function complexity gradually. For automatic verification, model checker for multi-agent systems (MCMAS) is adopted due to its novel capability to accommodate the multi-agent system and successfully verifies the targeting behaviours of the team-level autonomous systems. The verification results help retrospectively the design of decision-making algorithms improved by considering additional agents and behaviours during three steps of scenario modification. Consequently, the last scenario deals with the system composed of a ground control system, two unmanned aerial vehicles, and four unmanned ground vehicles with fault-tolerant and communication relay capabilities.     

一种双无人机协同跟踪地面移动目标方法

针对通信延时情况下双无人机协同跟踪地面移动目标问题进行研究, 构建了基于分布式遗传算法和滚动时域优化结合的目标跟踪航迹规划算法模型。考虑到通信延时会增加目标状态信息数据融合时的误差, 导致无人机跟踪任务效果变差, 结合递推最小二乘滤波和加权最小二乘估计设计了融合方法, 来融合处理目标状态信息; 考虑到无人机对目标的观测效果与未来时刻的目标状态信息密切相关, 采用递推最小二乘滤波预测目标的状态信息, 结合分布式遗传算法和滚动时域优化设计了双无人机目标跟踪航迹规划算法。适应度函数考虑了无人机和目标之间的距离、无人机之间的通信距离、无人机之间的通信角度。仿真结果表明：该协同跟踪方法能够较好地完成跟踪任务; 与一架无人机跟踪相比误差明显减小, 并且可以减小通信延时带来的跟踪误差。     

A Ground Control Station for a Multi-UAV Surveillance System

This paper presents the ground control station developed for a platform composed by multiple unmanned aerial vehicles for surveillance missions. The software application is fully based on open source libraries and it has been designed as a robust and decentralized system. It allows the operator to dynamically allocate different tasks to the UAVs and to show their operational information in a 3D realistic environment in real time. The ground control station has been designed to assist the operator in the challenging task of managing a system with multiple UAVs, trying to reduce his workload. The multi-UAV surveillance system has been demonstrated in field experiments using two quadrotors equipped with visual cameras.     

复杂环境下的交通标志检测与识别方法综述

交通标志检测与识别是无人驾驶三大模块中环境感知的研究热点之一,检测和识别交通标志可以向无人车传递道路交通信息,优化行车决策。在暴雨、大雾以及光线昏暗等复杂环境下,拍摄到的图像往往会被遮挡,变得模糊。这不仅影响图像的质量,还会对后期标志的检测与识别带来巨大的困难。简述了交通标志检测与识别方法,对近年来国内外学者解决各类复杂环境下交通标志检测与识别的方法、原理和步骤进行了总结归纳,有利于人们更好地解决此类问题。同时,对常用的交通标志数据集进行了总结,并对数据集里在复杂环境下拍摄的图像比例给予了说明。     

Modeling and Control of MM-UAV: Mobile Manipulating Unmanned Aerial Vehicle

Compared to autonomous ground vehicles, UAVs (unmanned aerial vehicles) have significant mobility advantages and the potential to operate in otherwise unreachable locations. Micro UAVs still suffer from one major drawback: they do not have the necessary payload capabilities to support high performance arms. This paper, however, investigates the key challenges in controlling a mobile manipulating UAV using a commercially available aircraft and a light-weight prototype 3-arm manipulator. Because of the overall instability of rotorcraft, we use a motion capture system to build an efficient autopilot. Our results indicate that we can accurately model and control our prototype system given significant disturbances when both moving the manipulators and interacting with the ground.     

Controlling Ensembles of Robots via a Supervisory Aerial Robot

We consider the task of controlling a large team of non-holonomic ground robots with an unmanned aerial vehicle in a decentralized manner that is invariant to the number of ground robots. The central idea is the development of an abstraction for the team of ground robots that allows the aerial platform to control the team without any knowledge of the specificity of individual vehicles. This happens in much the same way as a human operator can control a single robot vehicle by simply commanding the forward and turning velocities without a detailed knowledge of the specifics of the robot. The abstraction includes a gross model of the shape of the formation of the team, and information about the position and orientation of the team in the plane. We derive controllers that allow the team of robots to move in formation while avoiding collisions and respecting the abstraction commanded by the aerial platform. We propose strategies for controlling the physical spread of the ensemble of robots by splitting and merging the team based on distributed techniques. We provide simulation and experimental results using a team of indoor mobile robots and a three-dimensional, cable-controlled, parallel robot which serves as our indoor unmanned aerial platform.     

Radiation search operations using scene understanding with autonomous UAV and UGV

Autonomously searching for hazardous radiation sources requires the ability of the aerial and ground systems to understand the scene they are scouting. In this paper, we present systems, algorithms, and experiments to perform radiation search using unmanned aerial vehicles (UAV) and unmanned ground vehicles (UGV) by employing semantic scene segmentation. The aerial data are used to identify radiological points of interest, generate an orthophoto along with a digital elevation model (DEM) of the scene, and perform semantic segmentation to assign a category (e.g., road, grass) to each pixel in the orthophoto. We perform semantic segmentation by training a model on a dataset of images we collected and annotated, using the model to perform inference on images of the test area unseen to the model, and then refining the results with the DEM to better reason about category predictions at each pixel. We then use all of these outputs to plan a path for a UGV carrying a LiDAR to map the environment and avoid obstacles not present during the flight, and a radiation detector to collect more precise radiation measurements from the ground. Results of the analysis for each scenario tested favorably. We also note that our approach is general and has the potential to work for a variety of different sensing tasks.     

Decentralized cooperative control of heterogeneous vehicle groups

We coordinate in discrete time the interaction of two heterogeneous groups of mobile agents: a group of ground vehicles (ugvs) and a group of aerial vehicles (uavs). The ground agents interact with each other through time-invariant, nearest-neighbour rules. They synchronize their velocities through a specific communication protocol, and maintain cohesion and separation behaviour by means of interagent potential forces. Ground vehicles estimate their formation’s centroid using only locally available delayed information. That same information is transmitted to the aerial group, which orbits above the ground formation’s centroid, while avoiding midair collisions. Stability of the ground group motion is established in a Lyapunov framework. A Lyapunov analysis is also used to ensure that uavs track the ground group’s centroid.     

Autonomous Aerial Radio Repeating Using an A*-Based Path Planning Approach

In the event of a disaster, first responders must rapidly gain situational awareness about the environment in order to plan effective response operations. Unmanned ground vehicles are well suited for this task but often require a strong communication link to a remote ground station to effectively relay information. When considering an obstacle-rich environment, non-line-of-sight conditions and naive navigation strategies can cause substantial degradations in radio link quality. Therefore, this paper incorporates an unmanned aerial vehicle as a radio repeating node and presents a path planning strategy to cooperatively navigate the vehicle team so that radio link health is maintained. This navigation technique is formulated as an A*-based search and this paper presents the formulation of this path planner as well as an investigation into strategies that provide computational efficiency to the search process. The path planner uses predictions of radio signal health at different vehicle configurations to effectively navigate the vehicles and simulations have shown that the path planner produces favorable results in comparison to several conceivable naive radio repeating variants. The results also show that the radio repeating path planner has outperformed the naive variants in both simulated environments and in field testing where a Yamaha RMAX unmanned helicopter and a ground vehicle were used as the vehicle team.     

On the extraction of 3D models from airborne video sensors for geolocation

Geolocation of a feature in a video sequence collected from a moving platform is an activity that must be undertaken in video exploitation, especially surveillance and reconnaissance applications. Examples of sensor systems that are the focus of this work include manned and unmanned aerial vehicles. The approach described here uses positional information from three sources to compute refined coordinates in three dimensions for any feature in the video sequence. These three sources are: first, sensor-platform metadata describing the likely sensor footprint based on sensor-platform positional and attitudinal information; second, 3D information of a scene inherent in a video sequence collected from a moving platform; and third, reference imagery of the region of interest that is geolocated and georectified such as aerial photography. We describe the overall steps involved in this process and the progress made to date.     

Continuous Airborne Communication Relay Approach Using Unmanned Aerial Vehicles

As a result of unmanned aerial vehicles being widely used in different areas, studies about increasing the autonomous capabilities of unmanned aerial vehicles are gaining momentum. Today, unmanned aerial vehicle platforms are especially used in reconnaissance, surveillance and communications areas. In this study, in order to achieve continuous long-range communication relay infrastructure, artificial potential field based path planning of Unmanned Aerial Vehicles is discussed. A novel dynamic approach to relay-chain concept is proposed to maintain the communication between vehicles. Besides dynamically keeping vehicles in range and appropriate position to maintain communication relay, artificial potential field based path planning also provides collision avoidance system. The performance of the proposed system is measured by applying a simulation under the Matlab Simulink and Network Simulator environment. Artificial potential field based flight patterns are generated in Matlab, and performance of the communication between vehicles is measured in Network Simulation environment. Finally the simulation results show that an airborne communication relay can be established autonomously by using artificial potential filed based autonomous path planning approach. Continues state communication is provided by obtaining a resistant communication relay which depends on artificial potential field based positioning algorithm.     

基于图与势场法的多车道编队控制

多车协同驾驶能显著提高交通安全和效率,是未来5G网联自动驾驶技术的重要应用场景之一.传统上,多车协同驾驶的主要形式为单一车道上的无人车队列,其队列稳定性受队列长度、通信距离及延迟的限制.本文提出一种无人车编队方法,将单车道队列扩展为多车道护航编队.针对不同场景下的需求设计多车道编队调整策略,结合基于图的分布式控制,完成任意预定义的编队结构;同时,利用势场法对行车环境建立势场模型,实现无人车的避障轨迹规划,提高编队的避障能力;最后,结合纵横向控制器,实现无人车多车道护航编队控制.仿真实验表明,本文提出的无人车多车道护航编队方法,能适应不同交通场景,如道路变化、障碍车运动等,完成自动变换编队结构,实现安全、高效通行.     

无人机综合仿真与测试平台技术研究

固定翼无人机在军事,国防,民用上都具有非常广泛的应用。无人机系统包括三大部分：无人机机体,无人机地面控制站以及无人机地面综合检测站。详细介绍了当前已交付使用的一种适应于多种型号的无人机地面综合实验平台,包括无人机地面飞行仿真和地面检测和故障诊断三个部分。从整体设计方案出发,从技术角度分析各个子系统的功能和设计方案。提出了可以应用于无人机地面实验平台的基于专家系统的遥控-遥测知识对的知识库的建立和通过搜索知识库的故障诊断方法。并以一个平台子系统为例介绍了以任务管理和数据流控制为核心的软硬件设计方案。     

Moving observer trajectory control by angular measurements in tracking problem

An optimal path synthesis problem for a moving observer that performs angular observations over a target moving uniformly along a straight line on a plane is solved. It is supposed that elevation and azimuth angles can be observed when the observer moves in space and only the azimuth angle can be observed when the observer moves on a plane. Observer’s trajectories are obtained with the help of Pontryagin’smaximum principle as numerical solutions of an optimal control problem. As a performance criterion the trace of covariance matrix of the target motion elements estimate is used. A possibility of solving the problem in real time on board for unmanned aerial vehicle is investigated. A comparison with the scenario of two unmanned aerial vehicles using is given.     

非结构化环境中基于拓扑约束的地面无人驾驶路径规划算法

针对非结构化环境地面无人驾驶路径规划过程中路径避障以及多车路径冲突的难题,通过同调以及de Rham上同调对环境中障碍物拓扑信息的精确描述,提出一种拓扑约束下基于A*算法且用时更短的路径规划算法.该算法可实现非结构化环境中多无人车全局路径的拓扑分类,从而为多车的协同规划提供一种新的研究思路.此外,结合C-空间动态广义Voronoi图(GVD)的路径拓扑分离特性,提出一种拓扑约束下可用于多无人车全局路径规划的高效算法-----C-空间-GVD-${h_S     

Examining cognitive interference and adaptive safety behaviours in tactical vehicle control

Concurrent mental workload degrades some aspects of driving performance, but drivers might be able to modify their behaviour adaptively to accommodate cognitive impairments. For example, they might maintain longer vehicle headway in dual-task conditions to compensate for slowed response times. Studies documenting such adaptive behaviours typically use steady-state driving scenarios such as car following. Yet, driving often involves tactical control situations in which drivers need to monitor multiple aspects of their traffic environment and to accommodate changing goals. In two simulator experiments, this study examined the impact of mental workload on safety margins (distances) that drivers keep when engaged in a tactical control task: passing other vehicles. Although drivers did increase their headway adaptively when engaged in steady-state car following (experiment 2), they did not adapt their behaviour to accommodate cognitive load when performing tactical control manoeuvres. Implications of this difference between steady-state and tactical control driving contexts, both for driving research and for driving safety, are discussed.     

面向目标跟踪的多机协同通信保持控制

由于无人机存在通信和测量约束的情况,远程无人机执行持续目标跟踪任务时无法直接与地面站保持通信,需要其他无人机作为通信中继方可与地面站建立可靠的通信连接.基于Dubins曲线,采用最小转弯半径和航向调整相结合的方法对具有初始和终止航向角约束的多无人机进行协同航路规划,确保所有无人机同时到达指定位置,形成多机协同通信保持的初始构型.针对随机移动目标,在多机协同通信保持的动态过程中,考虑平台性能、通信约束、碰撞规避等约束条件,采用非线性模型预测控制(NMPC)实现无人机协同分布式在线优化.在确保无人机通信中继保持的前提下,有效提高了算法的实时性.仿真结果表明了该算法的有效性.     

Low-Cost Visual Tracking of a Landing Place and Hovering Flight Control with a Microcontroller

The growth of civil and military use has recently promoted the development of unmanned miniature aerial vehicles dedicated to surveillance tasks. These flying vehicles are often capable of carrying only a few dozen grammes of payload. To achieve autonomy for this kind of aircraft novel sensors are required, which need to cope with strictly limited onboard processing power. One of the key aspects in autonomous behaviour is target tracking. Our visual tracking approach differs from other methods by not using expensive cameras but a Wii remote camera, i.e. commodity consumer hardware. The system works without stationary sensors and all processing is done with an onboard microcontroller. The only assumptions are a good roll and pitch attitude estimation, provided by an inertial measurement unit and a stationary pattern of four infrared spots on the target or the landing spot. This paper details experiments for hovering above a landing place, but tracking a slowly moving target is also possible.