No description available.

The objective of this research is to demonstrate how an underlying system's state vector distribution can be determined in a distributed heterogeneous sensor network with reduced subspace observability at the individual nodes. We show how the network, as a whole, is capable of observing the target state vector even if the individual nodes are not capable of observing it locally. The initialization algorithm presented in this work can generate the initial state vector distribution for networks with a variety of sensor types as long as the measurements at the individual nodes are known functions of the target state vector. Initialization is accomplished through a novel distributed implementation of the particle filter that involves serial particle proposal and weighting strategies, which can be accomplished without sharing raw data between individual nodes in the network. The algorithm is capable of handling missed detections and clutter as well as compensating for delays introduced by processing, communication and finite signal propagation velocities. If multiple events of interest occur, their individual states can be initialized simultaneously without requiring explicit data association across nodes. The resulting distributions can be used to initialize a variety of distributed joint tracking algorithms. In such applications, the initialization algorithm can initialize additional target tracks as targets come and go during the operation of the system with multiple targets under track.

Under senare år har omfattande insatser gjorts för att främja mer hållbara flygtransporter i Europa. De konventionella tillverkningsmetoderna som används inom flyg- och rymdindustrin kräver betydande mängder råmaterial, vars utvinning, bearbetning och användning har negativa miljöeffekter. Därför finns det ett starkt incitament att utveckla nya, mer material-effektiva tillverkningsmetoder. Additiv tillverkning (AM), även känd som 3D-printining, har fördelen att direkt komma nära den slutliga formen på strukturer genom att lägga till material endast där det behövs, något som minimerar spill och förbättrar materialanvändningen. Dock utgör införandet av AM komponenter i säkerhetskritiska flyg- och rymdtillämpningar en betydande utmaning på grund av komplexiteten hos processerna. Denna komplexitet kan leda till tillverkningsvariationer som i sin tur kan resultera i defekter i de tillverkade strukturerna. Därför är framsteg inom automation genom utvecklingen av lösningar för övervakning och styrning under processens gång ett nödvändigt steg för att uppnå tillräcklig pålitlighet och repeterbarhet. Denna avhandling presenterar en utveckling av multisensorövervakning och styrning av Directed Energy Deposition (DED) med en laservärmekälla (LB). DED-LB är en avancerad teknik som möjliggör tillverkning av storskaliga metallkomponenter nära den slutliga formen. I detta arbete har lösningar undersökts för övervakning av DED-LB med tillsatspulver och tråd. För fallet med tillsatstråd kan denna kompletteras med resistiv förvärmning (så kallad hotwire), vilket ger möjlighet att ytterligare finjustera värmetillförseln och förbättra smältprocessen. För övervakningsändamål undersöktes tre olika in-situtekniker för processens stabilitet och varians. Maskinseende och elektriska givare användes för DED-LB med tillsatstråd (DED-LB/w), medan optisk spektroskopi användes för övervakning både av processen med tillsatspulver (DEDLB/p) samt med tråd. Ett multisensorsystem baserat på de tre teknologierna testades för DED-LB/w. Det kamerabaserade systemet gav tydliga indikationer på avvikelser från nominella processförhållanden. Spännings-och strömgivarnas signaler korrelerade med förändringar i processparametrar och återspeglade tydligt metallöverföringen. Spektrometersystemet indikerade förändringar relaterade till värmeöverföringen. Dessutom möjliggjorde analysen av erhållna spektra en detektering av förluster av viktiga legeringselement under DED-LB/p. Slutsatsen från resultaten understryker behovet av multisensorövervakning, eftersom det inte bara möjliggör detektering och skattning av processförändringar utan även en bättre förståelse av deras grundorsaker. Den presenterade ansatsen är ett viktigt bidrag i utvecklingen av ett framtida robust och feltolerant automatiskt styrsystem. / In recent years, an extensive effort has been made to leap European aviation towards more sustainable transportation. Conventional manufacturing methods used in aerospace industry require significant amounts of raw materials, whose extraction, processing, and utilization have adverse environmental impacts. Thus, there is a strong motivation to develop novel, more material efficient fabrication methods. Additive Manufacturing (AM), also known as 3D-printing, offers the advantage of manufacturing near-net-shape structures by adding material only where it is needed, minimizing waste, and improving material efficiency. However, introducing AM fabricated structures as components in safety-critical aerospace systems poses a significant challenge due to the inherent complexity of AM processes. This complexity can result in variations that may lead to defects or inconsistencies in the fabricated structures. Thus, increasing automation by developing in-process monitoring, and control solutions is the vital step to reach the necessary reliability and repeatability. This thesis presents development towards multi-sensor monitoring and control of Directed Energy Deposition (DED) using a Laser Beam (LB). DED-LB is an advanced technology that allows to manufacture large-scale, near-net-shape metallic parts. In this work, in-process monitoring solutions for DED-LB with feedstock powder and wire were investigated. The set-up of the latter was complemented by resistive pre-heating of the feedstock wire (hot-wire) which provided means of fine-tuning the heat input and improving metal fusion. Formonitoring purposes, three different in-situ techniques were investigated to monitor process stability and variability. Machine vision and electrical sensing were utilized during DED-LB with feedstock wire (DED-LB/w) depositions,while optical emission spectroscopy was used for monitoring processes with feedstock powder (DED-LB/p) as well as wire. A multi-sensorsystem based on the three sensing technologies was tested during DED-LB/w depositions. The vision system gave clear indications of variations from nominal conditions. Voltage and current sensors indications correlated to changes in process parameters and reflected well the metal transfer (liquid bridge) condition.The spectrometer system indicated well changes related to heat input. In addition, analysis of obtained spectra allowed to detect losses of vital alloying element during DED-LB/p. The main conclusion from the results underlines the need for simultaneous multi-sensor monitoring as it allows not only to detect and estimate process changes but also to better interpret their root causes. Such setup will positively enable a future robust, fault tolerant control system. / <p>Paper 3 is under acception but included in this thesis with CC BY-license.</p>

Data fusion in wireless sensor networks can improve the performance of a network by eliminating redundancy and power consumption, ensuring fault-tolerance between sensors, and managing e®ectively the available com- munication bandwidth between network components. This thesis considers a data fusion approach applied to wireless sensor networks based on fuzzy logic theory. In particular, a cluster-based hierarchical design in wire- less sensor networks is explored combined with two data fusion methods based on fuzzy logic theory. A data fusion algorithm is presented and tested using Mamdani and Tsukamoto fuzzy inference methods. In addition, a concept related to the appropriate queuing models is presented based on classical queuing theory. Results show that the Mamdani method gives better results than the Tsukamoto approach for the two implementations considered. We noted that the proposed algorithm requires low processing and computational power. As a result, it can be applied to WSNs to provide optimal data fusion and ensures maximum sensor lifetime and minimum time delay.

Le vieillissement du parc aéronautique, mais aussi la volonté de prolonger le cycle de vie des appareils, impose aux techniques de maintenance des exigences de contrôle accrues en termes de fiabilité et de rapidité. Les principaux enjeux étant la détection, et surtout la caractérisation des micro-fissures pouvant apparaitre dans les pièces sensibles des appareils. Cette caractérisation pouvant aller jusqu'à la reconstruction qualitative, voire quantitative du profil des fissures, ce qui suppose la résolution du problème inverse consistant, à partir des signaux mesurés, à remonter à certaines caractéristiques de la pièce inspectée, notamment géométriques, qui en sont en partie la cause.Ce manuscrit présente une approche originale de conception de sonde d'imagerie par courants de Foucault. Cette approche, ici appliquée à l'imagerie de pièces aéronautiques cylindriques, consiste à concevoir la sonde de façon à satisfaire la double contrainte d'une instrumentation performante et d'une résolution possible du problème inverse. Ainsi, la conception de la sonde est-elle faite de manière à ce que d'une part les signaux mesurés, résultant de ses interactions avec la pièce contrôlée, soient d'amplitude la plus élevée possible, et d'autre part, a n que ces interactions soient modélisables au moyen d'un modèle qui se prête bien à l'inversion.Dans un premier temps, une méthodologie de conception de sonde est présentée, qui vise à optimiser le dimensionnement et l'agencement de ses éléments en se fixant des contraintes de sensibilité et de résolution. Un prototype, conçu d'après cette méthodologie, est réalisé, mis en œuvre et caractérisé. Les résultats expérimentaux obtenus, qui s'avèrent en accord avec une modélisation par éléments finis, offrent une validation du fonctionnement de la sonde, celle-ci permettant effectivement de mettre en évidence les défauts types (micro-fissures) recherchés dans la pièce inspectée. Dans un second temps, la configuration des interactions sonde-pièce inspectée, choisie à dessein, est mise à profit pour élaborer un modèle direct. Celui-ci repose sur l'hypothèse selon laquelle, les défauts étant de faibles dimensions, il est possible de considérer que la perturbation de signal qu'ils engendrent est équivalente à celle que génèreraient des sources virtuelles de courants localisées uniquement dans le volume du défaut. Outre sa simplicité, cette modélisation offre l'avantage d'être adaptée à une mise en œuvre au moyen de la méthode des points sources distribués. Or celle-ci permet de formuler le problème direct sous forme matricielle ce qui constitue une base de nature à faciliter la résolution du problème inverse.C'est à ce problème qu'est consacrée la dernière partie du mémoire. Des méthodes d'inversion y sont proposées, visant à traiter progressivement la complexité du problème. Ainsi un algorithme d'inversion mono-fréquence est-il proposé, qui se montre efficace pour reconstruire des défauts sur de faibles profondeurs. Pour une meilleure reconstruction en profondeur, des algorithmes multifréquence faisant appel à des méthodes de régularisation sont ensuite conçus et appliqués à des signaux correspondant à des défauts de géométries diverses. / The ageing of the aeronautical fleet and the will to increase the aircrafts lifetime require the maintenance techniques to be made always more reliable and fast. In this context, the detection and characterization of the microscopic cracks likely to appear in some sensitive parts of the aircrafts is an important issue to be faced.This work deals with an original approach for the design of an eddy current imaging probe dedicated to the non destructive evaluation of cylindrical fastener parts. This approach consists in designing the probe in such a way that it both satisfies the constraint of an efficient sensing and that of enabling the inverse problem to be solved with a view to the defects reconstruction.Firstly, a probe design methodology is presented that optimizes the emission/reception topology, the choice of the probe elements and their arrangement according to sensitivity and resolution constraints. A prototype is built, implemented and characterized and validation of the design is obtained as the researched defects are displayed with experimental performances that agree with finite elements modelling simulations.Secondly, since the probe relies on a uniform eddy current flow interacting with small defects, a rather simple forward model is proposed based on virtual defect current sources (VDCS). The model implementation is carried out using the distributed point source method leading to a matrix formulation that can facilitate the resolution of the inverse problem.Finally, mono-frequency and multi-frequency methods are proposed for inverting the VDCS forward model and promising defect reconstruction results are obtained.

Approved for public release; distribution in unlimited. / Different sensors exploit different regions of the electromagnetic spectrum; therefore, a multi-sensor image fusion system can take full advantage of the complementary capabilities of individual sensors in the suit; to produce information that cannot be obtained by viewing the images separately. In this thesis, a framework for the multiresolution fusion of the night vision devices and thermal infrared imagery is presented. It encompasses a wavelet-based approach that supports both pixel-level and region-based fusion, and aims to maximize scene content by incorporating spectral information from both the source images. In pixel-level fusion, source images are decomposed into different scales, and salient directional features are extracted and selectively fused together by comparing the corresponding wavelet coefficients. To increase the degree of subject relevance in the fusion process, a region-based approach which uses a multiresolution segmentation algorithm to partition the image domain at different scales is proposed. The region's characteristics are then determined and used to guide the fusion process. The experimental results obtained demonstrate the feasibility of the approach. Potential applications of this development include improvements in night piloting (navigation and target discrimination), law enforcement etc. / Civilian, Republic of Singapore

The focus of this research is concerned with performing forecasting in technoeconomic contexts using a set of certain novel artificial neural networks (ANNs). Relevant efforts in general, entail the task of quantitatively estimating the details about the likelihood of future events (or unknown outcomes/effects) based on past and current information on the observed events (or known causes). Commensurate with the scope and objectives of the research, the specific topics addressed are as follows: A review on various methods adopted in technoeconomic forecasting and identified are econometric projections that can be used for forecasting via artificial neural network (ANN)-based simulations Developing and testing a compatible version of ANN designed to support a dynamic sigmoidal (squashing) function that morphs to the stochastical trends of the ANN input. As such, the network architecture gets pruned for reduced complexity across the span of iterative training schedule leading to the realization of a constructive artificial neural-network (CANN). Formulating a training schedule on an ANN with sparsely-sampled data via sparsity removal with cardinality enhancement procedure (through Nyquist sampling) and invoking statistical bootstrapping technique of resampling applied on the cardinality-improved subset so as to obtain an enhanced number of pseudoreplicates required as an adequate ensemble for robust training of the test ANN: The training and prediction exercises on the test ANN corresponds to optimally elucidating output predictions in the context of the technoeconomics framework of the power generation considered Prescribing a cone-of-error to alleviate over- or under-predictions toward prudently interpreting the results obtained; and, squeezing the cone-of-error to get a final cone-of-forecast rendering the forecast estimation/inference to be more precise Designing an ANN-based fuzzy inference engine (FIE) to ascertain the ex ante forecast details based on sparse sets of ex post data gathered in technoeconomic contexts - Involved thereof a novel method of .fusing fuzzy considerations and data sparsity.Lastly, summarizing the results with essential conclusions and identifying possible research items for future efforts identified as open-questions. / Includes bibliography. / Dissertation (Ph.D.)--Florida Atlantic University, 2014. / FAU Electronic Theses and Dissertations Collection

An ocean turbine extarcts the kinetic energy from ocean currents to generate electricity. Machine Condition Monitoring (MCM) / Prognostic Health Monitoring (PHM) systems allow for self-checking and automated fault detection, and are integral in the construction of a highly reliable ocean turbine. MCM/PHM systems enable real time health assessment, prognostics and advisory generation by interpreting data from sensors installed on the machine being monitored. To effectively utilize sensor readings for determining the health of individual components, macro-components and the overall system, these measurements must somehow be combined or integrated to form a holistic picture. The process used to perform this combination is called data fusion. Data mining and machine learning techniques allow for the analysis of these sensor signals, any maintenance history and other available information (like expert knowledge) to automate decision making and other such processes within MCM/PHM systems. ... This dissertation proposes an MCM/PHM software architecture employing those techniques which were determined from the experiments to be ideal for this application. Our work also offers a data fusion framework applicable to ocean machinery MCM/PHM. Finally, it presents a software tool for monitoring ocean turbines and other submerged vessels, implemented according to industry standards. / by Janell A. Duhaney. / Thesis (Ph.D.)--Florida Atlantic University, 2012. / Includes bibliography. / Mode of access: World Wide Web. / System requirements: Adobe Reader.

Os três principais aspectos da robótica móvel são: mapear ambientes, conhecer a postura do robô nesse ambiente e controlá-lo de forma a atingir um objetivo. Possuindo um mapa e conhecendo-se a postura inicial do robô neste mapa é possível saber a nova postura através das informações do sistema de odometria do robô a cada instante de tempo. Em função de diferenças no solo, derrapagens das rodas, dentre outros fatores, a odometria gera um erro que se acumula com o tempo (drift). Este trabalho aborda a investigação de um procedimento baseado em filtragem de Kalman para estimação de postura do robô em ambientes in-door utilizando sensores simples computacionalmente (odômetro e scanner a laser) objetivando-se eliminar o problema de drift. Resultados experimentais baseados em ambiente real mostram o bom desempenho do procedimento, mesmo com ambiente bastante ruidoso.

A fusão de dados ée uma disciplina relativamente nova que procura combinar dados de diversos sensores para realizar inferências que não seriam possíveis com um único sensor apenas. Este trabalho traz uma implementação conceitual de um sistema especialista aplicado a um cenário de vigilância aérea com o objetivo de auxiliar na tomada de decisão em diversas situações. A arquitetura proposta envolve lógica fuzzy como ferramenta para inferência, podendo ser classificada como fusão de dados nível dois. A vantagem desta escolha ée a capacidade de se combinar grandezas lingüísticas, ou simbólicas, com dados quantitativos. A motivação para este trabalho foi que a reação do sistema especialista a uma determinada situação pode ser mais rápida que a humana e ainda englobar uma maior quantidade de informações disponíveis, tanto de sensores como de níveis inferiores de fusão de dados. Desta forma, o sistema apresentado torna-se de grande auxílio na tomada de decisão.