Recognition of a landing platform for unmanned aerial vehicles by using computer vision-based techniques

The use of Unmanned Aerial Vehicles (UAVs) is growing significantly for many and varied purposes. During the mission, an outdoor UAV is guided by following the planned path using GPS signals. However, the GPS capability may become defective or the environment may be GPS-denied, and an additional safety aid is therefore required for the automatic landing phase that is independent of GPS data. Most UAVs are equipped with machine vision systems which, together with onboard analysis, can be used for safe, automatic landing. This contributes greatly to the overall success of autonomous flight.This paper proposes an automatic expert system, based on image segmentation procedures, that assists safe landing through recognition and relative orientation of the UAV and platform. The proposed expert system exploits the human experience that has been incorporated into the machine vision system, which is mapped into the proposed image processing modules. The result is an improved reliability capability that could be incorporated into any UAV, and is especially robust for rotary wing UAVs. This is clearly a desirable fail-safe capability.     

Guidance and nonlinear control system for autonomous flight of minirotorcraft unmanned aerial vehicles

Small unmanned aerial vehicles (UAVs) are becoming popular among researchers and vital platforms for several autonomous mission systems. In this paper, we present the design and development of a miniature autonomous rotorcraft weighing less than 700 g and capable of waypoint navigation, trajectory tracking, visual navigation, precise hovering, and automatic takeoff and landing. In an effort to make advanced autonomous behaviors available to mini‐ and microrotorcraft, an embedded and inexpensive autopilot was developed. To compensate for the weaknesses of the low‐cost equipment, we put our efforts into designing a reliable model‐based nonlinear controller that uses an inner‐loop outer‐loop control scheme. The developed flight controller considers the system's nonlinearities, guarantees the stability of the closed‐loop system, and results in a practical controller that is easy to implement and to tune. In addition to controller design and stability analysis, the paper provides information about the overall control architecture and the UAV system integration, including guidance laws, navigation algorithms, control system implementation, and autopilot hardware. The guidance, navigation, and control (GN&C) algorithms were implemented on a miniature quadrotor UAV that has undergone an extensive program of flight tests, resulting in various flight behaviors under autonomous control from takeoff to landing. Experimental results that demonstrate the operation of the GN&C algorithms and the capabilities of our autonomous micro air vehicle are presented. © 2009 Wiley Periodicals, Inc.     

量子机器学习综述

近年来, 机器学习一直是被关注和探讨的研究热点, 被应用到各领域并在其中起着重要作用. 但随着数据量的不断增加, 机器学习算法训练时间越来越长. 与此同时, 量子计算机表现出强大的运算能力. 因此, 有研究人员尝试用量子计算的方法解决机器学习训练时间长的问题, 量子机器学习这一领域应运而生. 量子主成分分析、量子支持向量机、量子深度学习等量子机器学习算法相继被提出, 并有实验证明了量子机器学习算法有显著的加速效果, 使得量子机器学习的研究展现出逐步走高的趋势. 对量子机器学习算法进行综述. 首先介绍量子计算基础; 然后对量子监督学习、量子无监督学习、量子半监督学习、量子强化学习以及量子深度学习5类量子机器学习算法进行介绍; 接着对量子机器学习的相关应用进行介绍并给出了算法实验; 最后进行总结和展望.     

UAV Task Assignment

Unmanned aerial vehicles (UAVs) are becoming vital warfare and homeland security platforms because they have the potential to significantly reduce cost and risk to human life while amplifying warfighter and first-responder capabilities. This article builds on the very active area of planning and control for autonomous multiagent systems. This work represents a step toward enabling robust decision making for distributed autonomous UAVs by improving the team's operational reliability and capabilities through better system self-awareness and adaptive mission planning. The health-aware task assignment algorithm developed in this article was demonstrated to be effective both in simulation and flight experiments.     

Deep Imitation Learning for Autonomous Vehicles Based on Convolutional Neural Networks

Providing autonomous systems with an effective quantity and quality of information from a desired task is challenging. In particular, autonomous vehicles, must have a reliable vision of their workspace to robustly accomplish driving functions. Speaking of machine vision, deep learning techniques, and specifically convolutional neural networks, have been proven to be the state of the art technology in the field. As these networks typically involve millions of parameters and elements, designing an optimal architecture for deep learning structures is a difficult task which is globally under investigation by researchers. This study experimentally evaluates the impact of three major architectural properties of convolutional networks, including the number of layers, filters, and filter size on their performance. In this study, several models with different properties are developed,equally trained, and then applied to an autonomous car in a realistic simulation environment. A new ensemble approach is also proposed to calculate and update weights for the models regarding their mean squared error values. Based on design properties,performance results are reported and compared for further investigations. Surprisingly, the number of filters itself does not largely affect the performance efficiency. As a result, proper allocation of filters with different kernel sizes through the layers introduces a considerable improvement in the performance.Achievements of this study will provide the researchers with a clear clue and direction in designing optimal network architectures for deep learning purposes.     

Application of reinforcement learning for agent-based production scheduling

Reinforcement learning (RL) has received some attention in recent years from agent-based researchers because it deals with the problem of how an autonomous agent can learn to select proper actions for achieving its goals through interacting with its environment. Although there have been several successful examples demonstrating the usefulness of RL, its application to manufacturing systems has not been fully explored yet. In this paper, Q-learning, a popular RL algorithm, is applied to a single machine dispatching rule selection problem. This paper investigates the application potential of Q-learning, a widely used RL algorithm to a dispatching rule selection problem on a single machine to determine if it can be used to enable a single machine agent to learn commonly accepted dispatching rules for three example cases in which the best dispatching rules have been previously defined. This study provided encouraging results that show the potential of RL for application to agent-based production scheduling.     

Multi-UAV Simulator Utilizing X-Plane

This paper describes the development of a simulator for multiple Unmanned Aerial Vehicles (UAVs) utilizing the commercially available simulator X-Plane and Matlab. Coordinated control of unmanned systems is currently being researched for a wide range of applications, including search and rescue, convoy protection, and building clearing to name a few. Although coordination and control of Unmanned Ground Vehicles (UGVs) has been a heavily researched area, the extension towards controlling multiple UAVs has seen minimal attention. This lack of development is due to numerous issues including the difficulty in realistically modeling and simulating multiple UAVs. This work attempts to overcome these limitations by creating an environment that can simultaneously simulate multiple air vehicles as well as provide state data and control input for the individual vehicles using a heavily developed and commercially available flight simulator (X-Plane). This framework will allow researchers to study multi-UAV control algorithms using realistic unmanned and manned aircraft models in real-world modeled environments. Validation of the system’s ability is shown through the demonstration of formation control algorithms implemented on four UAV helicopters with formation and navigation controllers built in Matlab/Simulink.     

Development and application of an autonomous unmanned aerial vehicle for precise aerobiological sampling above agricultural fields

Remote‐controlled (RC) unmanned aerial vehicles (UAVs) have been used to study the movement of agricultural threat agents (e.g., plant and animal pathogens, invasive weeds, and exotic insects) above crop fields, but these RC UAVs are operated entirely by a ground‐based pilot and often demonstrate large fluctuations in sampling height, sampling pattern, and sampling speed. In this paper, we describe the development and application of an autonomous UAV for precise aerobiological sampling tens to hundreds of meters above agricultural fields. We equipped a Senior Telemaster UAV with four aerobiological sampling devices and a MicroPilot‐based autonomous system, and we conducted 25 sampling flights for potential agricultural threat agents at Virginia Tech's Kentland Farm. To determine the most appropriate sampling path for aerobiological sampling above crop fields with an autonomous UAV, we explored five different sampling patterns, including multiple global positioning system (GPS) waypoints plotted over a variety of spatial scales. An orbital sampling pattern around a single GPS waypoint exhibited high positional accuracy and produced altitude standard deviations ranging from 1.6 to 2.8 m. Autonomous UAVs have the potential to extend the range of aerobiological sampling, improve positional accuracy of sampling paths, and enable coordinated flight with multiple UAVs sampling at different altitudes. © 2008 Wiley Periodicals, Inc.     

Tracking control of trim trajectories of a blimp for ascent and descent flight manoeuvres

A blimp is a small airship that has no metal framework and collapses when deflated. It belongs to family of unmanned aerial vehicles (UAVs). In this paper we address the problem of designing tracking feedback control of an underactuated autonomous UAV. The ascent and descent flight conditions as one in which the rate of change (of magnitude) of the airship's state vector is zero and the resultant of the applied forces and moments is constant lead to trimmed flight trajectories. The subject of the tracking control is to stabilize the engine around the planned flight. Using a combined integrator backstepping approach and Lyapunov theory, the stability results are local and overcome the minimum number of actuators (inputs) with respect to the blimp's six degrees of freedom. Considering physic limits in UAVs, other trimmed flights are investigated and compared.     

A systematic literature review of flying ad hoc networks: State-of-the-art,challenges, and perspectives

Unmanned aerial vehicles (UAVs), also known as drones, communicate, collaborate, and form flying ad hoc networks (FANETs) to perform many different missions, ranging from delivery tasks to agriculture applications. Recently, FANETs have been integrated with different technologies, such as artificial intelligence (AI), virtual reality, and Internet of Things. Such new avenues for the use of UAVs directly impact the research on FANETs and cause some major challenges, such as security and physical layer issues, resource management, and UAV positioning issues that need to be addressed. Several researchers have been working for the last few years to propose AI and machine learning (ML)-based solutions for different use cases in UAV-based networks. They present the limitations of the existing research work and highlight some possible future works on FANETs. However, exhibiting the trends in the UAV research papers in a quantitative manner is still required to motivate researchers to rethink the research on FANETs. Therefore, this study covers more than 170 scientific publications extracted from five trusted academic databases published from 2013 to 2021 to provide a thorough overview of the main research and development statistics in the area of FANETs, the open challenges existing in this area and the ML-based solutions to solve these challenges. In addition, the investigation of emerging technologies integrated with FANETs, as well as the simulation tools employed for evaluating FANETs' performance are discussed. Moreover, the future research directions in the area of FANETs are considered within a prospective vision discussion.     

Indoor Mobile Robotics at Grima, PUC

This paper describes the main activities and achievements of our research group on Machine Intelligence and Robotics (Grima) at the Computer Science Department, Pontificia Universidad Catolica de Chile (PUC). Since 2002, we have been developing an active research in the area of indoor autonomous social robots. Our main focus has been the cognitive side of Robotics, where we have developed algorithms for autonomous navigation using wheeled robots, scene recognition using vision and 3D range sensors, and social behaviors using Markov Decision Processes, among others. As a distinguishing feature, in our research we have followed a probabilistic approach, deeply rooted in machine learning and Bayesian statistical techniques. Among our main achievements are an increasing list of publications in main Robotics conference and journals, and the consolidation of a research group with more than 25 people among full-time professors, visiting researchers, and graduate students.     

Robotic technologies for solar‐powered UAVs: Fully autonomous updraft‐aware aerial sensing for multiday search‐and‐rescue missions

Large‐scale aerial sensing missions can greatly benefit from the perpetual endurance capability provided by high‐performance low‐altitude solar‐powered unmanned aerial vehicles (UAVs). However, today these UAVs suffer from small payload capacity, low energetic margins, and high operational complexity. To tackle these problems, this paper presents four individual technical contributions and integrates them into an existing solar‐powered UAV system: First, a lightweight and power‐efficient day/night‐capable sensing system is discussed. Second, means to optimize the UAV platform to the specific payload and to thereby achieve sufficient energetic margins for day/night flight with payload are presented. Third, existing autonomous launch and landing functionality is extended for solar‐powered UAVs. Fourth, as a main contribution an extended Kalman filter (EKF)‐based autonomous thermal updraft tracking framework is developed. Its novelty is that it allows the end‐to‐end integration of the thermal‐induced roll moment into the estimation process. It is assessed against unscented Kalman filter and particle filter methods in simulation and implemented on the aircraft's low‐power autopilot. The complete system is verified during a 26 h search‐and‐rescue aerial sensing mock‐up mission that represents the first‐ever fully autonomous perpetual endurance flight of a small solar‐powered UAV with a day/night‐capable sensing payload. It also represents the first time that solar‐electric propulsion and autonomous thermal updraft tracking are combined in flight. In contrast to previous work that has focused on the energetic feasibility of perpetual flight, the individual technical contributions of this paper are considered core functionality to guarantee ease‐of‐use, effectivity, and reliability in future multiday aerial sensing operations with small solar‐powered UAVs.     

A binocular vision-based UAVs autonomous aerial refueling platform

Unmanned aerial vehicles (UAVs) are highly focused and widely used in various domains, and the capability of autonomous aerial refueling (AAR) becomes increasingly important. Most of the research in this area concerns the verification of the algorithms while the experiments are conducted on the ground. In this work, in order to verify the vision system designed for boom approach AAR, an integrated platform is built and tested. The platform consists of a tanker UAV, a receiver UAV and a ground station. The pictures of the marker on the receiver UAV are captured by the binocular vision system on the tanker UAV and then used for flight control and boom control. Performance and feasibility of the platform are demonstrated by the real out-door flight tests, and the experimental results verified the feasibility and effectiveness of our developed binocular vision-based UAVs AAR.     

Cooperative Control of UAVs for Localization of Intermittently Emitting Mobile Targets

Compared with a single platform, cooperative autonomous unmanned aerial vehicles (UAVs) offer efficiency and robustness in performing complex tasks. Focusing on ground mobile targets that intermittently emit radio frequency signals, this paper presents a decentralized control architecture for multiple UAVs, equipped only with rudimentary sensors, to search, detect, and locate targets over large areas. The proposed architecture has in its core a decision logic which governs the state of operation for each UAV based on sensor readings and communicated data. To support the findings, extensive simulation results are presented, focusing primarily on two success measures that the UAVs seek to minimize: overall time to search for a group of targets and the final target localization error achieved. The results of the simulations have provided support for hardware flight tests.      

Unmanned Aerial Vehicles: Control Methods and Future Challenges

With the rapid development of computer technology,automatic control technology and communication technology,research on unmanned aerial vehicles(UAVs)has attracted extensive attention from all over the world during the last decades.Particularly due to the demand of various civil applications,the conceptual design of UAV and autonomous flight control technology have been promoted and developed mutually.This paper is devoted to providing a brief review of the UAV control issues,including motion equations,various classical and advanced control approaches.The basic ideas,applicable conditions,advantages and disadvantages of these control approaches are illustrated and discussed.Some challenging topics and future research directions are raised.     

Sigmoid Limiting Functions and Potential Field Based Autonomous Air Refueling Path Planning for UAVs

In this paper autonomous air-refueling (AAR) path planning for Unmanned Aerial Vehicles (UAVs) has been discussed and an enhanced approach has been put forward. AAR path planning for UAVs was designed and the basic model of the pattern was put forward in our previous work (Cetin and Yilmaz 2013). Additionally to our previous works, the deficiencies of the previous approach, like smooth maneuvers in the tanker approach and the boundary functions of the potential zones has been handled, furthermore special pattern parameters are added to the approach which makes it suitable for different kind of UAVs that has variable flight speed and turn radius parameters. An important originality of the approach is using of sigmoid limiting functions while modeling dynamic behaviors of the potential fields that are based on path planning algorithms. In order to use the AAR path planning approach in a real time application, the computation is performed in Graphical Processing Units (GPUs) based parallel architecture by benefiting from many cores in General Purpose Graphical Processing Units (GPGPU) as described in previous research (Cetin and Yilmaz 2013). With the addition of the sigmoid limiting functions instead of logical binary boundary functions computation needs of the autonomous approach become higher point and the only way to use the approach in the real time applications is benefiting of the parallel computing approach. The comparison of the boundary functions as computational performance and path outputs are discussed with the simulation results in this paper. Simulation results are proved that this novel autonomous parallel path planning approach is successful and it would be used in real time applications like AAR mission.     

机器学习模型在车险欺诈检测的研究进展

随着保险行业的蓬勃发展,保险欺诈问题也显得日趋严重。车险欺诈一直是保险欺诈的“重灾区”,对保险行业的发展至关重要。因此,车险欺诈检测技术一直是国内外学者研究的热点问题。鉴于我国在机动车辆保险欺诈检测技术方相对滞后,而国外的研究成果又较少对我国车险业务数据进行有效建模与分析,首次针对机器学习模型应用在车险欺诈检测的研究工作进行了文献调研,对二十多年来的研究工作进行系统化的归纳与总结。通过引入车险欺诈流程的简介,对专家系统与智能理赔系统在车险欺诈检测的流程进行了叙述;依次从国外和国内的角度介绍了机器学习模型应用在车险欺诈检测的具体研究进展,并进行了宏观的对比;基于国内某车险公司提供近5年来高质量的车险数据选取具有代表性的机器学习模型进行建模,并进行了全面的测试与分析;探讨了车险欺诈检测技术未来的研究方向。     

Fuzzy Logic Based Approach to Design of Autonomous Landing System for Unmanned Aerial Vehicles

This paper is concerned with autonomous flight of UAVs and proposes a fuzzy logic based autonomous flight and landing system controller. Besides three fuzzy logic controllers which are developed for autonomous navigation for UAVs in a previous work as fuzzy logic based autonomous mission control blocks, three more fuzzy logic modules are developed under the main landing system for the control of the horizontal and the vertical positions of the aircraft against the runway under a TACAN (Tactical Air Navigation) approach. The performance of the fuzzy logic based controllers is evaluated using the standard configuration of MATLAB and the Aerosim Aeronautical Simulation Block Set which provides a complete set of tools for rapid development of 6 degree-of-freedom nonlinear generic manned/unmanned aerial vehicle models. Additionally, FlightGear Flight Simulator and GMS aircraft instruments are deployed in order to get visual outputs that aid the designer in evaluating the performance and the potential of the controllers. The simulated test flights on an Aerosonde indicate the capability of the approach in achieving the desired performance despite the simple design procedure.     

GPS‐denied Indoor and Outdoor Monocular Vision Aided Navigation and Control of Unmanned Aircraft

GPS‐denied closed‐loop autonomous control of unstable Unmanned Aerial Vehicles (UAVs) such as rotorcraft using information from a monocular camera has been an open problem. Most proposed Vision aided Inertial Navigation Systems (V‐INSs) have been too computationally intensive or do not have sufficient integrity for closed‐loop flight. We provide an affirmative answer to the question of whether V‐INSs can be used to sustain prolonged real‐world GPS‐denied flight by presenting a V‐INS that is validated through autonomous flight‐tests over prolonged closed‐loop dynamic operation in both indoor and outdoor GPS‐denied environments with two rotorcraft unmanned aircraft systems (UASs). The architecture efficiently combines visual feature information from a monocular camera with measurements from inertial sensors. Inertial measurements are used to predict frame‐to‐frame transition of online selected feature locations, and the difference between predicted and observed feature locations is used to bind in real‐time the inertial measurement unit drift, estimate its bias, and account for initial misalignment errors. A novel algorithm to manage a library of features online is presented that can add or remove features based on a measure of relative confidence in each feature location. The resulting V‐INS is sufficiently efficient and reliable to enable real‐time implementation on resource‐constrained aerial vehicles. The presented algorithms are validated on multiple platforms in real‐world conditions: through a 16‐min flight test, including an autonomous landing, of a 66 kg rotorcraft UAV operating in an unconctrolled outdoor environment without using GPS and through a Micro‐UAV operating in a cluttered, unmapped, and gusty indoor environment. © 2013 Wiley Periodicals, Inc.     

Coordinated aerobiological sampling of a plant pathogen in the lower atmosphere using two autonomous unmanned aerial vehicles

Unmanned aerial vehicles (UAVs) are an important tool to track the long‐distance movement of plant pathogens above crop fields. Here, we describe the use of a control strategy (coordination via speed modulation) to synchronize two autonomous UAVs during aerobiological sampling of the potato late blight pathogen, Phytophthora infestans. The UAVs shared position coordinates via a wireless mesh network and modulated their speeds so that they were properly phased within their sampling orbits. Three coordinated control experiments were performed August 14–15, 2008. In the first two experiments, two UAVs were vertically aligned at two different altitudes [25 and 45 m above ground level (AGL)] with identical sampling orbits (radii of 150 m). In the third experiment, two UAVs shared the same altitude (35 m AGL) with different sampling orbits (radii of 130 and 160 m). Orbit times did not vary significantly between the two UAVs across all three aerobiological sampling missions, and the phase error during sampling converged to zero within 2 min following the start of the coordinated control algorithm. Viable sporangia of P. infestans were recovered following two of the coordinated flights. This is the first detailed report of autonomous UAV coordination during the aerobiological sampling of a plant pathogen in the lower atmosphere. UAVs operating independently of one another may experience significant sampling variations during the course of a flight. Coordinating the flight of two or more UAVs ensures that the vehicles enter, sample, and exit a spore plume at consistent times. © 2010 Wiley Periodicals, Inc.