On Autonomous Target Tracking for UAVs PDF Download

Author: Panagiotis Theodorakopoulos
Publisher:
ISBN:
Category :
Languages : en
Pages : 150

View

Book Description
Most applications of Unmanned Aerial Vehicles are related to events that occur on the ground. In particular, ground target tracking, be the target static, slowly moving or maneuvering at high speeds, is an essential task for UAVs. The overall objective of this thesis is to provide methods to endow a drone to autonomously track a moving ground target, under the following conditions: - A fixed wing UAV equipped with a monocular camera. - Presence of obstacles that hinder ground visibility. - No Fly Zones that limit the airspace. - Restrictions on the field of view of the observing sensor (a camera) - Various target dynamics and behavior: the target may be either moving on an open field or on a road network, and also has dynamic constraints (e.g. if it is a car). It can be neutral or evasive: in the latter case, it can exploit the presence of obstacles, denoted as "shadows" to avoid being tracked by the UAV, making the problem akin to a "hide and seek" game. The thesis proposes three approaches to tackle this problem: - A control based navigation method, - An adversarial predictive method, - And a discrete game theoretic approach. Results obtained in realistic simulations and with an actual UAV are presented to evaluate and compare the pros and cons of each approach. Extensions to the multi-drone case are also considered.

Author: Joshua Patrick Effland
Publisher:
ISBN:
Category : Computer vision
Languages : en
Pages :

View

Book Description

Author: T. M. Lam
Publisher: IntechOpen
ISBN: 9789537619411
Category : Technology & Engineering
Languages : en
Pages : 780

View

Book Description
This book contains 35 chapters written by experts in developing techniques for making aerial vehicles more intelligent, more reliable, more flexible in use, and safer in operation.It will also serve as an inspiration for further improvement of the design and application of aeral vehicles. The advanced techniques and research described here may also be applicable to other high-tech areas such as robotics, avionics, vetronics, and space.

Author: Kimon P. Valavanis
Publisher: Springer Science & Business Media
ISBN: 1402061145
Category : Technology & Engineering
Languages : en
Pages : 552

View

Book Description
The past decade has seen tremendous interest in the production and refinement of unmanned aerial vehicles, both fixed-wing, such as airplanes and rotary-wing, such as helicopters and vertical takeoff and landing vehicles. This book provides a diversified survey of research and development on small and miniature unmanned aerial vehicles of both fixed and rotary wing designs. From historical background to proposed new applications, this is the most comprehensive reference yet.

Author: Randal W. Beard
Publisher: Princeton University Press
ISBN: 1400840600
Category : Technology & Engineering
Languages : en
Pages : 317

View

Book Description
Autonomous unmanned air vehicles (UAVs) are critical to current and future military, civil, and commercial operations. Despite their importance, no previous textbook has accessibly introduced UAVs to students in the engineering, computer, and science disciplines--until now. Small Unmanned Aircraft provides a concise but comprehensive description of the key concepts and technologies underlying the dynamics, control, and guidance of fixed-wing unmanned aircraft, and enables all students with an introductory-level background in controls or robotics to enter this exciting and important area. The authors explore the essential underlying physics and sensors of UAV problems, including low-level autopilot for stability and higher-level autopilot functions of path planning. The textbook leads the student from rigid-body dynamics through aerodynamics, stability augmentation, and state estimation using onboard sensors, to maneuvering through obstacles. To facilitate understanding, the authors have replaced traditional homework assignments with a simulation project using the MATLAB/Simulink environment. Students begin by modeling rigid-body dynamics, then add aerodynamics and sensor models. They develop low-level autopilot code, extended Kalman filters for state estimation, path-following routines, and high-level path-planning algorithms. The final chapter of the book focuses on UAV guidance using machine vision. Designed for advanced undergraduate or graduate students in engineering or the sciences, this book offers a bridge to the aerodynamics and control of UAV flight.

Author: Ruoyu Tan
Publisher:
ISBN:
Category :
Languages : en
Pages : 95

View

Book Description
An Unmanned Air Vehicle (UAV) is an aircraft without a human pilot on board. It can be controlled either autonomously by computers onboard, or using a remote control by a pilot on the ground, or in another vehicle. In both military and civilian sectors, UAVs are quickly obtaining popularity and expected to expand dramatically in the years to come. As UAVs gain more attention, one of the immediate requirements would be to have UAVs work as much autonomously as possible. One of the common tasks that UAVs would be engaged in is target tracking which has various potential applications in military field, law-enforcement, wildlife protection effort, and so on. This thesis focuses on development of a controller for UAVs to track ground target. In particular, this thesis focuses on quadrotor UAV, which is a multicopter that is lifted and propelled using four motors. Admittedly, several target tracking control methods have been developed in recent years. However, only a few of them have been applied on a quadrotor. Most of these tracking methods, particularly those based on Proportional Derivative (PD) control laws, which have been applied on quadrotors, are not time efficient due to practical acceleration constraint and a number of parameters that need to be tuned. The UAV control problem can be divided into 4 sub-problems: Position Control, Motor Control, Trajectory Tracking and Trajectory Generation. In this thesis, the dynamic equations of motion for quadrotors and a Proportional Derivative control law is derived to solve the problems of Position Control, Motor Control and Trajectory Tracking. A Proportional Navigation (PN) based switching strategy is proposed to address the problem of Trajectory Generation. The experiments and numerical simulations are performed using non-maneuvering and maneuvering targets. The simulation results show that the proposed PN based switching strategy not only carries out effective tracking but also results into smaller oscillations and errors when compared to the widely used PD tracking method. The switching strategy, as proposed as a solution to target tracking problem, leaves an important question with regard to when should the switching happen. It is intuitive that the time of switching will play a role in how fast the UAV converges to the target. The second problem considered in this thesis relates to the optimal time of switching that would minimize the positional error between the UAV and the target. An optimal switching strategy is proposed to obtain the optimal switching time for both non-maneuvering and maneuvering targets. Analytical solutions that generate trajectories based on PN and PD methods are used in this strategy. The numerical simulations validate the optimality, reliability, and accuracy of the proposed method for both non-maneuvering and maneuvering targets.

Author: Mohit Sewak
Publisher: Springer
ISBN: 9811382859
Category : Computers
Languages : en
Pages : 203

View

Book Description
This book starts by presenting the basics of reinforcement learning using highly intuitive and easy-to-understand examples and applications, and then introduces the cutting-edge research advances that make reinforcement learning capable of out-performing most state-of-art systems, and even humans in a number of applications. The book not only equips readers with an understanding of multiple advanced and innovative algorithms, but also prepares them to implement systems such as those created by Google Deep Mind in actual code. This book is intended for readers who want to both understand and apply advanced concepts in a field that combines the best of two worlds – deep learning and reinforcement learning – to tap the potential of ‘advanced artificial intelligence’ for creating real-world applications and game-winning algorithms.

Author: Mohsen Majnoon
Publisher:
ISBN:
Category :
Languages : en
Pages : 0

View

Book Description
Control of unmanned aerial vehicles is a very active topic in research with lots of applications ranging from civilian to military. To control a UAV, its attitude is often controlled using gyroscopes, but to control its position, inertial sensors together with GPS are often used. However, obtaining accurate current position is difficult using inertial sensors because of the integration drift. GPS on the other hand is not functional in indoor applications since it cannot connect to GPS satellites. Since vision has been proved to be an inexpensive and consistent source of relative position information, vision-based control is getting more popular in UAVs recently, but then again, using vision in outdoor applications is challenging as the target can move fast and out of the vision sensor field of view. So, in order to keep the target inside the field of view, two algorithms are being developed and tested via simulation in this research. Using pan/tilt/zoom cameras or multi camera systems, the target is guaranteed to stay in vision system field of view and hence, the vision based pose estimation can provide the control system with proper relative position. Two case studies - vision-based mobile-target tracking of a quadrotor using a multi-camera vision sensor and vision-based mobile-target tracking of a tilting rotor aircraft equipped with a zooming camera - are presented in this research to show the applicability of these methods in UAV control.

Author: Hao Dong
Publisher: Springer Nature
ISBN: 9811540950
Category : Computers
Languages : en
Pages : 526

View

Book Description
Deep reinforcement learning (DRL) is the combination of reinforcement learning (RL) and deep learning. It has been able to solve a wide range of complex decision-making tasks that were previously out of reach for a machine, and famously contributed to the success of AlphaGo. Furthermore, it opens up numerous new applications in domains such as healthcare, robotics, smart grids and finance. Divided into three main parts, this book provides a comprehensive and self-contained introduction to DRL. The first part introduces the foundations of deep learning, reinforcement learning (RL) and widely used deep RL methods and discusses their implementation. The second part covers selected DRL research topics, which are useful for those wanting to specialize in DRL research. To help readers gain a deep understanding of DRL and quickly apply the techniques in practice, the third part presents mass applications, such as the intelligent transportation system and learning to run, with detailed explanations. The book is intended for computer science students, both undergraduate and postgraduate, who would like to learn DRL from scratch, practice its implementation, and explore the research topics. It also appeals to engineers and practitioners who do not have strong machine learning background, but want to quickly understand how DRL works and use the techniques in their applications.

Author: Venkata Ramaiah Gidda
Publisher:
ISBN:
Category :
Languages : en
Pages : 117

View

Book Description
In recent years, surveillance and border patrol have become one of the key research areas in UAV research. Increase in the computational capability of the computers and embedded electronics, coupled with compatibility of various commercial vision algorithms and commercial off the shelf (COTS) embedded electronics, and has further fuelled the research. The basic task in these applications is perception of environment through the available visual sensors like camera. Visual tracking, as the name implies, is tracking of objects using a camera. The process of autonomous target tracking starts with the selection of the target in a sequence of video frames transmitted from the on-board camera. We use an improved fast dynamic template matching algorithm coupled with Kalman Filter to track the selected target in consecutive video frames. The selected target is saved as a reference template. On the ground station computer, the reference template is overlaid on the live streaming video from the on-board system, starting from the upper left corner of the video frame. The template is slid pixel by pixel over the entire source image. A comparison of the pixels is performed between the template and source image. A confidence value R of the match is calculated at each pixel. Based on the method used to perform the template matching, the best match pixel location is found according to the highest or lowest confidence value R. The best match pixel location is communicated to the on-board gimbal controller over the wireless Xbee network. The software on the controller actuates the pan-tilt servos to continuously to hold the selected target at the center of the video frame. The complete system is a portable control system assembled from commercial off the shelf parts. The tracking system is tested on a target having several motion patterns.