The problem considered in this thesis is the development of an autonomous prototype robot capable of gathering sensory information from its environment allowing it to provide feedback on the condition of specific targets to aid in maintenance of hydro equipment. The context for the solution to this problem is based on the power grid environment operated by the local hydro utility. The intent is to monitor power line structures by travelling along skywire located at the top of towers, providing a view of everything beneath it including, for example, insulators, conductors, and towers. The contribution of this thesis is a novel robot design with the potential to prevent hazardous situations and the use of rough coverage feedback modified reinforcement learning algorithms to establish behaviours. / October 2007

This dissertation describes the design and implementation of a system that has been designed to extract 3D information from pairs of 2D images. System input consists of two images taken by an ordinary digital camera. System output is a full 3D model extracted from 2D images. There are no assumptions about the positions of the cameras during the time when the images are being taken, but the scene must not undergo any modifications. The process of extracting 3D information from 2D images consists of three basic steps. First, point matching is performed. The main contribution of this step is the introduction of an approach to matching image segments in the context of an approximation space. The second step copes with the problem of estimating external camera parameters. The proposed solution to this problem uses 3D geometry rather than the fundamental matrix widely used in 2D to 3D conversion. In the proposed approach (DirectGS), the distances between reprojected rays for all image points are minimised. The contribution of the approach considered in this step is a definition of an optimal search space for solving the 2D to 3D conversion problem and introduction of an efficient algorithm that minimises reprojection error. In the third step, the problem of dense matching is considered. The contribution of this step is the introduction of a proposed approach to dense matching of 3D object structures that utilises the presence of points on lines in 3D space. The theory and experiments developed for this dissertation demonstrate the usefulness of the proposed system in the process of digitizing 3D information. The main advantage of the proposed approach is its low cost, simplicity in use for an untrained user and the high precision of reconstructed objects. / October 2007

This thesis presents a rough set approach to reinforcement learning. This is made possible by considering behaviour patterns of learning agents in the context of approximation spaces. Rough set theory introduced by Zdzisław Pawlak in the early 1980s provides a ground for deriving pattern-based rewards within approximation spaces. Learning can be considered episodic. The framework provided by an approximation space makes it possible to derive pattern-based reference rewards at the end of each episode. Reference rewards provide a standard for reinforcement comparison as well as the actor-critic method of reinforcement learning. In addition, approximation spaces provide a basis for deriving episodic weights that provide a basis for a new form of off-policy Monte Carlo learning control method. A number of conventional and pattern-based reinforcement learning methods are investigated in this thesis. In addition, this thesis introduces two learning environments used to compare the algorithms. The first is a Monocular Vision System used to track a moving target. The second is an artificial ecosystem testbed that makes it possible to study swarm behaviour by collections of biologically-inspired bots. The simulated ecosystem has an ethological basis inspired by the work of Niko Tinbergen, who introduced in the 1960s methods of observing and explaining the behaviour of biological organisms that carry over into the study of the behaviour of interacting robotic devices that cooperate to survive and to carry out highly specialized tasks. Agent behaviour during each episode is recorded in a decision table called an ethogram, which records features such as states, proximate causes, responses (actions), action preferences, rewards and decisions (actions chosen and actions rejected). At all times an agent follows a policy that maps perceived states of the environment to actions. The goal of the learning algorithms is to find an optimal policy in a non-stationary environment. The results of the learning experiments with seven forms of reinforcement learning are given. The contribution of this thesis is a comprehensive introduction to a pattern-based evaluation of behaviour during reinforcement learning using approximation spaces. / May 2006

En esta memoria se incluyen diversas alternativas a una teoría de krein-milman no arquimediana las cuales vienen sugeridas bien por intentos anteriores de otros autores bien por conseguir una teoría unificada en los casos arquimediano o no o bien por lograr una teoría independiente del cuerpo valuado y que en condiciones de comparación dan lugar a resultados muy similares / This monography provides several alternatives to a non-Archimedan Krein-Milman Theory. They are suggested by some previous attempts to this subject carried out by other authors, as well as by the aim of getting an unified theory that works in the Archimede and in the non-Archimedean cases, in the sense that in the Archimedean context, this theory coincides with the well-known one existing in the classical literature

Fonaments de la Matemàtica per al tractament de la Incertesa. Noves aportacions a l’estudi de les Equacions Borroses i de les Equacions Diferencials Borroses. Aplicacions de la Matemàtica de la Incertesa al comportament de models de la teoria econòmica.

To simplify troubleshooting and reliability of a process, a diagnosis system can supervise the process and alarm if any faults are detected. A diagnosis system can also identify one, or several faults, i.e. isolate faults, that may have caused the alarm. If model-based diagnosis is used, tests based on observations from the process are compared to a model of the process to diagnose the process. It can be a hard task to find which tests to be used for maximal fault detection and fault isolation. Structural Methods require not very detailed knowledge of the process to be diagnosed and can be used to find such tests early in the design of new processes. Sensors are used to get observations of a process. Therefore, sensors placed on different positions in the process gives different possibilities for observations. A specific set of sensors are in this work called a sensor configuration. This thesis contributes with a method to predict and examine the fault detection and fault isolation possibility. By using these two diagnosis properties, a suitable sensor configuration is computed and tests to be used in a future diagnosis system are suggested. For this task an algorithm which can be used in the design phase of diagnosis systems, and a Matlab implementation of this algorithm are described. In one part of this work the Matlab implementation and the algorithm are used to study how a model-based diagnosis-system can be used to supervise the fuel system in an Unmanned Aerial Vehicle (UAV).

Individual differences on complex memory spans predict a variety of higher-order cognitive tasks (e.g. reading comprehension, reasoning, following direction) as well as low-level attention tasks (e.g. Stroop, dichotic listening, antisaccade). The current study attempted to better determine the role of individual differences in complex memory span and episodic retrieval. Specifically, two experiments explored the possibility that individual differences in complex memory span reflect differences in the ability to successfully retrieve items from secondary memory via a cue-dependent search process. High and low complex span participants were tested in delayed (Experiment 1) and continuous distractor (Experiment 2) free recall with varying list-lengths. Across both experiments low spans recalled fewer items than high spans, recalled more previous list intrusions than high spans, and recalled at a slower rate than high spans. It is argued that low spans search through a larger set of items than high spans and, thus low spans episodic retrieval deficits are associated with an inability to use cues to guide a search and retrieval process of secondary memory. Implications for dual-component models of memory are discussed.

The probability hypothesis density (PHD), popularized by Ronald Mahler, presents a novel and theoretically-rigorous approach to multitarget, multisensor tracking. Based on random set theory, the PHD is the first moment of a point process of a random track set, and it can be propagated by Bayesian prediction and observation equations to form a multitarget, multisensor tracking filter. The advantage of the PHD filter lies in its ability to estimate automatically the expected number of targets present, to fuse easily different kinds of data observations, and to locate targets without performing any explicit report-to-track association. We apply a particle-filter implementation of the PHD filter to realistic multitarget, multisensor tracking using passive coherent location (PCL) systems that exploit illuminators of opportunity such as FM radio stations. The objective of this dissertation is to enhance the usefulness of the PHD particle filter for multitarget, multisensor tracking, in general, and within the context of PCL, in particular. This involves a number of thrusts, including: 1) devising intelligent proposal densities for particle placement, 2) devising a peak-extraction algorithm for extracting information from the PHD, 3) incorporating realistic probabilities of detection and signal-to-noise ratios (including multipath effects) to model realistic PCL scenarios, 4) using range, Doppler, and direction of arrival (DOA) observations to test the target detection and data fusion capabilities of the PHD filter, and 5) clarifying the concepts behind FISST and the PHD to make them more accessible to the practicing engineer. A goal of this dissertation is to serve as a tutorial for anyone interested in becoming familiar with the probability hypothesis density and associated PHD particle filter. It is hoped that, after reading this thesis, the reader will have gained a clearer understanding of the PHD and the functionality and effectiveness of the PHD particle filter.

Magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) has been developed and applied to clinical analysis studies due to its non-invasive properties. Because of the increasing clinical interests of applying MRS, a lot of post-processing tools has been developed, among which LCModel is one of the most popular. LCModel estimates the absolute metabolite concentrations in vivo according to the basis file, so basis files play an important role for the accuracy of absolute metabolite concentrations. The default basis sets of LCModel were made by phantom experiments. However, some special metabolites are difficult to get them, so the basis sets lack for these metabolites. In order to avoid this trouble, LCModel provides a special method called ¡§spectra offering¡¨. In this study, we use GAMMA Visual Analysis (GAVA) software to create basis sets and compare the shape of LCModel default basis sets with the shape of GAVA basis sets. Some metabolites which are not included in the LCModel phantom experiments are also generated. Finally, we estimate the absolute concentrations in normal subjects and patients by using these two kinds of basis sets respectively. Using LCModel ¡§spectra offering¡¨ method to append extra metabolites for LCModel basis sets is applicable to those metabolites of singlet resonance but not of J-coupling resonance in the meanwhile. Our results demonstrate that using GAVA simulation as the basis set leads to different quantitative results from using basis sets of in vitro. We believe that using GAVA simulation as the basis set would provide better consistency among all metabolites and thus achieve more accurate quantification of MRS.

This thesis aims comparative analysis of the state of the art shot boundary detection algorithms. The major methods that have been used for shot boundary detection such as pixel intensity based, histogram-based, edge-based, and motion vectors based, are implemented and analyzed. A recent method which utilizes &ldquo / graph partition model&rdquo / together with the support vector machine classifier as a shot boundary detection algorithm is also implemented and analyzed. Moreover, a novel graph theoretic concept, &ldquo / dominant sets&rdquo / , is also successfully applied to the shot boundary detection problem as a contribution to the solution domain.