This thesis investigates the possibility of using reinforcement learning (RL) techniques to create a flight controller for a quadrotor Micro Aerial Vehicle (MAV). A capable flight control system is a core requirement of any unmanned aerial vehicle. The challenging and diverse applications in which MAVs are destined to be used, mean that considerable time and effort need to be put into designing and commissioning suitable flight controllers. It is proposed that reinforcement learning, a subset of machine learning, could be used to address some of the practical difficulties. While much research has delved into RL in unmanned aerial vehicle applications, this work has tended to ignore low level motion control, or been concerned only in off-line learning regimes. This thesis addresses an area in which accessible information is scarce: the performance of RL when used for on-policy motion control. Trying out a candidate algorithm on a real MAV is a simple but expensive proposition. In place of such an approach, this research details the development of a suitable simulator environment, in which a prototype controller might be evaluated. Then inquiry then proposes a possible RL-based control system, utilising the Q-learning algorithm, with an adaptive RBF-network providing function approximation. The operation of this prototypical control system is then tested in detail, to determine both the absolute level of performance which can be expected, and the effect which tuning critical parameters of the algorithm has on the functioning of the controller. Performance is compared against a conventional PID controller to maximise the usability of the results by a wide audience. Testing considers behaviour in the presence of disturbances, and run-time changes in plant dynamics. Results show that given sufficient learning opportunity, a RL-based control system performs as well as a simple PID controller. However, unstable behaviour during learning is an issue for future analysis. Additionally, preliminary testing is performed to evaluate the feasibility of implementing RL algorithms in an embedded computing environment, as a general requirement for a MAV flight controller. Whilst the algorithm runs successfully in an embedded context, observation reveals further development would be necessary to reduce computation time to a level where a controller was able to update sufficiently quickly for a real-time motion control application. In summary, the study provides a critical assessment of the feasibility of using RL algorithms for motion control tasks, such as MAV flight control. Advantages which merit interest are exposed, though practical considerations suggest at this stage, that such a control system is not a realistic proposition. There is a discussion of avenues which may uncover possibilities to surmount these challenges. This investigation will prove useful for engineers interested in the opportunities which reinforcement learning techniques represent.

Small Unmanned Aerial Vehicles (SUAVs) are an attractive choice for many surveillance tasks. However, video from an SUAV can be difficult to use in its raw form. In addition, the limitations inherent in the SUAV platform inhibit the distribution of video to remote users. To solve the problems with using SUAV video, we propose a system to automatically create geo-referenced mosiacs of video frames. We also present three novel techniques we have developed to improve ortho-rectification and geo-location accuracy of the mosaics. The most successful of these techniques is able to reduce geo-location error by a factor of 15 with minimal computational overhead. The proposed system overcomes communications limitations by transmitting the mosaics to a central server where there they can easily be accessed by remote users via the Internet. Using flight test results, we show that the proposed mosaicking system achieves real-time performance and produces high-quality and accurately geo-referenced imagery.

Miniature Aerial Vehicles (MAVs) are quickly gaining acceptance as a platform for performing remote sensing or surveillance of remote areas. However, because MAVs are typically flown close to the ground (1000 feet or less in altitude), their field of view for any one image is relatively small. In addition, the context of the video (where and at what orientation are the objects being observed, the relationship between images) is unclear from any one image. To overcome these problems, we propose a geo-referenced mosaicing method that creates a mosaic from the captured images and geo-references the mosaic using information from the MAV IMU/GPS unit. Our method utilizes bundle adjustment within a constrained optimization framework and topology refinement. Using real MAV video, we have demonstrated our mosaic creation process on over 900 frames. Our method has been shown to produce the high quality mosaics to within 7m using tightly synchronized MAV telemetry data and to within 30m using only GPS information (i.e. no roll and pitch information).

Published Article / Today's industries are based on an automated workplace. These automated workplaces are efficient, reconfigurable and intelligent automated environments. They are filled with technology, robotics, Automated Guided Vehicle (AGV) and, or Unmanned Aerial Vehicles (UAV) etc. For full automation will one need to effectively track an object, unmanned aerial vehicle (UAV) or automated guided vehicle (AGV). Effective tracking of vehicles can be used for control. This could result in less hardware on the craft that leads to a longer battery life, a bigger pay load or more processing power. This system track by using a stationary colour camera placed at an optimal placing in the automated workplace. The vehicle or objects are painted in two colours (colour A and colour B) that are not present in the automated workplace. The images from the camera are hue colour filtered to extract only the object or vehicle. The area, placement in frame and relationship between colour A and B are used for position and determine the orientation of AGV, UAV or object.

published_or_final_version / Geography and Geology / Doctoral / Doctor of Philosophy

No description available.

No description available.

There has been recent interest by the military to have platforms capable of operating close to a point of interest without being detected while providing critical surveillance. By providing information that is not readily available, these platforms could provide a useful tool for small unit commanders in potentially life-threatening situations. Highly maneuverable, slow-flying micro air vehicles could fly under canopies, through alleys, or indoors to provide such intelligence. This study consists of a survey of current micro-sized technologies and commercially available components. The findings are presented and used in the design process of a larger payload-carrying variant of the NPS flapping wing micro air vehicle. The intent is to develop a readily deployable, backpackable, slow-flying micro air vehicle that can be used by smaller-size ground units in theatre for urban reconnaissance.

Our related research review on propagation models reveals six factors that are significant in last mile connectivity via LAP: path loss, elevation angle, LAP altitude, coverage area, power consumption, operation frequency, interference, and antenna type. These factors can help with monitoring system performance, network planning, coverage footprint, receivers' line-of-sight, quality of service requirements, and data rates which may all vary in response to geomorphology characteristics. Several competing propagation models have been proposed over the years but whilst they collectively raise many shortcomings such as limited altitude up to few tens of meters, lack of cover across different environments, low perdition accuracy they also exhibit several advantages. Four propagation models, which are representatives of their types, have been selected since they exhibit advantages in relation to high altitude, wide coverage range, adaption across different terrains. In addition, all four have been extensively deployed in the past and as a result their correction factors have evolved over the years to yield extremely accurate results which makes the development and evaluation aspects of this research very precise. The four models are: ITU-R P.529-3, Okumura, Hata-Davidson, and ATG. The aim of this doctoral research is to design a new propagation model for last-mile connectivity using LAPs technology as an alternative to aerial base station that includes all six factors but does not exhibit any of the shortcomings of existing models. The new propagation model evolves from existing models using machine learning. The four models are first adapted to include the elevation angle alongside the multiple-input multiple-output diversity gain, our first novelty in propagation modelling. The four adapted models are then used as input in a Neural Network framework and their parameters are clustered in a Self-Organizing-Map using a minimax technique. The framework evolves an optimal propagation model that represents the main research contribution of this research. The optimal propagation model is deployed in two proof-of-concept applications, a wireless sensor network, and a cellular structure. The performance of the optimal model is evaluated and then validated against that of the four adapted models first in relation to predictions reported in the literature and then in the context of the two proof-of-concept applications. The predictions of the optimised model are significantly improved in comparison to those of the four adapted propagation models. Each of the two proof-of-concept applications also represent a research novelty.

No description available.