Made available in DSpace on 2014-06-12T15:52:23Z (GMT). No. of bitstreams: 1 license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2009 / Devido aos problemas relacinados com o gerenciamento e o raciocínio em grandes ontologias, um forte interesse e uma conseqüente investigação ativa em modularização de ontologias vem emergindo na comunidade científica relacionada à Web Semântica. Uma vez que muitas ontologias são preferivelmente grandes artefatos, para uma adoção em grande escala das mesmas, por exemplo, na Web Semântica, é necessário permitir que desenvolvedores de ontologias possam incluir apenas as entidades e axiomas que são relevantes para a aplicação que estão a desenvolver. Além da reutilização, o uso de modularização de ontologias é útil para muitas outras tarefas, incluindo suporte a queries, raciocínio distribuído, desenvolvimento em grande escala e manutenção de ontologias. Alguns abordagens para modularização têm sido propostas, no entanto nemhuma delas dispõe de uma ferramenta flexível que permita não só definição de módulos, mas também outros tipos de tarefas, como análise sintática e semântica, biblioteca de módulos e uma poderosa ferramenta para seleção de entidades. Esta dissertação propõe um conjunto de ferramentas, chamado ModOnto, para cumprir esses requisitos. Incorpora uma abordagem para modularização de ontologias que herda alguns dos princípios de Engenharia de Software Orientada para Objeto, que são o encapsulamento e o ocultamento de informação

Thesis deals with support of educational courses focused to single chip microprocessors in secondary schools. It includes a text introducing students to the use of single microprocessor and provides examples to their applications.

This study investigated the relationship between the laboratory environments and the learning styles of middle school technology education teachers in the Commonwealth of Virginia. Based on the assumption that a strong relationship between teaching and learning styles exists, it was hypothesized that teacher preference for one type of laboratory over another (conventional or modular) may be an issue of learning style. A random sample (n=195) was drawn from the entire population (as identified by the Virginia Department of Education in 1998) of public middle school technology education teachers (N=392). Randomly selected teachers were mailed a cover letter, demographic questionnaire, postage-paid return envelope, the Learning Type Measure (LTM) instrument, and one dollar for taking the time to complete and return the instrument. The LTM instrument, demographic questionnaire and Bernice McCarthy's research on the 4MAT System of Leadership and Instruction were used to describe the laboratory environments and the teaching and learning styles of the respondents. Data collected were compared using contingency tables and Pearson's Chi-square analysis. Eighty-three (42.5%) of the middle school teachers responded and sixty-five of the instruments (78%) were usable. The findings indicate that respondents were overwhelmingly male (94%) and had considerable teaching experience (mean = 17.4). Sixty-percent of respondents taught in a modular laboratory and forty-percent taught in a conventional laboratory. Of the four learning styles identified by the LTM (Imaginative, Analytic, Common Sense, and Dynamic), respondents overwhelmingly (69.2%) rated themselves as Common Sense learners. Common Sense learners as teachers encourage practical applications, are interested in productivity and competence, like technical things, use hands-on activities, and try to give students the skills they will need to be economically independent in life. These findings are consistent with previous research involving the personalities and learning styles of industrial arts/technology educators. The self-perceived learning styles of respondents were significantly different when compared to McCarthy's findings for secondary teachers and administrators in general. However, the learning styles of respondents in conventional laboratories were not significantly different than the learning styles of respondents in modular laboratories. Though it seems logical that learning style might explain laboratory preference, this notion was not supported by this study. / Ed. D.

International Telemetering Conference Proceedings / October 18-21, 2004 / Town & Country Resort, San Diego, California / The function of a PCM telemetry encoder, installed in moving vehicles such as automobiles, aircraft, missiles, and artillery projectiles, is to transform many physical variables, such as velocity, shock, temperature, vibration and pressure, into digital data. Also, the encoder is required to make a data frame composed of digital input signals and frame synchronous data. The framed data is supplied to the input of a transmitter. There are three critical considerations in developing a PCM telemetry encoder to be installed in an artillery projectile. The first is the performance consideration, such as sampling rate, data receiving rate and data transmission rate. The second is the size consideration due to the severely limited installation space in an artillery projectile and the last is the power consumption consideration due to limitations of the munition’s power supply. To meet these three considerations, the best alternative is a one-chip solution. Using a commercially available TMS320F2812 DSP chip, we have implemented a 30-channel PCM telemetry encoder to process randomized data frames, composed of 16-channel analog data, 14-channel digital data and 2 frame synchronization data per data frame, at 10Mbps transmission baud rate. This paper describes the structure of the 30-channel PCM telemetry encoder and its performance.

This research investigates tolerance strategies for modular systems on a project specific basis. The objective of the proposed research is to form a guideline for optimizing the construction costs/risks with the aim of developing an optimal design of resilient modular systems. The procedures for achieving the research objective included: (a) development of 3D structural analysis models of the modules, (b) strength/stability investigation of the structure, (c) developing the fabrication cost function, (e) checking elastic and inelastic distortion, and (f) constructing the site-fit risk functions. The total site-fit risk function minimizes the cost/risk associated with fabrication, transportation; alignment, rework, and safety, while maximizing stiffness in terms of story drift values for site re-alignment and fitting alternatives. The fabrication cost function was developed by collecting 61 data points for the investigated module chassis using the SAP2000 software while reducing the initial section sizes, in addition to the fabrication costs at each step (61 steps). With the reduction of the structural reinforcement, story drift values increase, therefore there will be a larger distortion in the module. This generic module design procedure models a trade-off between the amount of reinforcement and expected need for significant field alterations. Structural design software packages such as SAP2000, AutoCAD, and Autodesk were used in order to model and test the module chassis. This research hypothesizes that the influential factors in the site-fit risk functions are respectively: fabrication, transportation, alignment, safety, and rework costs/risks. In addition, the site-fit risk function provides a theoretical range of possible solutions for the construction industry. The maximum allowable modular out-of-tolerance value, which requires the minimum amount of cost with respect to the defined function, can be configured using this methodology. This research concludes that over-reinforced or lightly-reinforced designs are not the best solution for mitigating risks, and reducing costs. For this reason the site-fit risk function will provide a range of pareto-optimal building solutions with respect to the fabrication, transportation, safety, alignment, and rework costs/risks.

Title: Compact objects in categories of modules Author: Peter Kálnai Department: Department of Algebra Supervisor: Mgr. Jan Žemlička, Ph.D., Department of Algebra Abstract: In the thesis we state baic properties of compact objects in various appropriate categories like categories of modules, stable factor category over a perfect ring and Grothendieck categories. We find a ring R such that the class of dually slender R-modules is closed under direct products under some set-theoretic assumption. Finally, we characterize the conditions, when countably generat- ed projective modules are finitely generated, expressed by their Grothendieck monoid. Keywords: compact, dually slender module, stable module category, projective module, self-small

Le présent travail de thèse se propose de profiter de la description cristallographique à N dimensions pour rediscuter les structures des alliages métalliques périodiques dont les atomes sont situés sur un sous-ensemble ordonné d’un Z-module — c’est-à-dire sur des positions qui sont des combinaisons linéaires entières de N>3 vecteurs arithmétiquement indépendants— exemplifiés par les quasicristaux et les phases approximantes. La description cristallographique à N dimensions permet ainsi de révéler des symétries supplémentaires, cachées dans l’espace physique, susceptibles d’engendrer de nouveaux types de défauts, dits défauts de module, dans les alliages métalliques périodiques dont les atomes se localisent sur un Z-module non trivial dans l’espace direct. Ces défauts correspondent aux opérations de symétrie interne du Z-module qui sont perdues lors de la projection rationnelle dans l’espace physique en raison de la périodicité de l’alliage. Ce sont des macles, des défauts de translations et les dislocations qui les bordent et que nous désignons par dislocations de modules pour les différencier des usuelles dislocations de réseau. Elles apparaissent comme des dislocations partielles bordant une ou plusieurs fautes d’empilement. Des dislocations particulières peuvent exister avec des vecteurs de Burgers ayant une composante nulle dans l’espace physique. Ces dislocations, que nous appelons dislocations scalaires, n’engendrent aucun champ de contrainte et ne sont sensibles à aucun champ de déformation. La phase orthorhombique NiZr de symétrie Cmcm a une maille cristalline dont tous les atomes sont situés sur une fraction d’un même Z-module pentagonal ; elle, est à ce titre, parfaitement descriptible à 5 dimensions. A parti de cette analyse, nous avons prédit l’existence de nouveaux défauts tels que les macles quinaires et défauts de translations, tous observés et confirmés par nos observations en microscopie électronique haute résolution HREM et STEM-HAADF. / The present work is an attempt to take advantage of the N-dimensional crystallographic description to rediscuss the structures of periodic metal alloys whose atoms are located on an ordered subset of a Z-module — positions that are integer linear combinations of N> 3 arithmetically independent vectors — like those encountered in quasicrystals and approximants. This N-dimensional crystallographic description makes it possible to reveal hidden symmetries in periodic metallic alloys, that can generate new types of possible defects by symmetry breaking during the projection in the physical space. Thus, in addition to the usual defects due to a local breaking of periodicity, these alloys may contain new defects corresponding to internal symmetry operations of the Z-module which are lost due to the periodicity of the alloy. These defects are twins, translation defects and dislocations that border them and that we refer to as dislocations of modules to differentiate them from the usual network dislocations and which appear as partial dislocations bordering one or more stacking faults. Particular dislocations may exist with Burgers vectors having a null component in the physical space. These dislocations, which we call scalar dislocations, generate no stress field and are not sensitive to any deformation field. The orthorhombic phase of NiZr has a crystal structure the atoms of which are located on a fraction of the pentagonal Z-module and is therefore fully describable at 5 dimensions. We can thus describe new defects such as quinary twins, but also defects of translations that all keep the Z-module invariant. All these defects have indeed been observed and recognized by high resolution electron microscopy (HREM and STEM-HAADF).

This research applies a recently developed dynamic state-estimation based protection scheme, the settingless protection, to the photovoltaic (PV) industry for the first time. At this time, the proposed protection algorithm has been implemented on traditional protection zones for individual power system devices, but this research extends this protection to a microgrid, specifically, a system of PV network composed of several PV modules. Several illustrative examples on various anomalies such as high impedance faults and shorted-out PV modules have been provided to demonstrate the effectiveness of this protection scheme. The detection of these anomalies has been demonstrated in the presence of changing atmospheric conditions, and with the operation of maximum power point tracking (MPPT) equipped dc-dc converters. This protection scheme requires an accurate model of the PV module, therefore, a two-diode PV model has been developed using quadratic integration modeling. In this PV model development, a scaling factor is applied to the Taylor series expansion of the exponential terms of the model of the PV module. Then the higher order terms of the Taylor series expansion are reduced to at most second order terms using the quadratization technique. Furthermore, a novel approach for extracting the PV parameters, namely, the ideality constants, leakage currents, PV module internal current, shunt and series resistances, has been presented. A comparison was performed between numerically generated data using the determined PV module parameters and data measurements from a physical PV module. It was shown that the maximum error from this comparison was below 0.12A, and less than 0.05A around the maximum power point region of the PV modules used for this research. The residual data from the PV array protection scheme has been used to develop a method for identifying the location of faulted PV modules. Also, condition-based monitoring of the PV array system has also been presented with examples. From the PV array system monitoring, the shading and underperformance of a PV module have been identified. From the contributions of this research, an accurate module of the PV array has been developed in a form that can be integrated with other power system devices. This accurate module can be used for state estimation of the PV array, load flow analysis, short circuit analysis, and other power system analytical studies. Also, by determining the location of the faulted PV module, the time to identify this faulted PV module in a large PV installation is drastically reduced. Lastly, by identifying shading conditions and underperforming PV modules, the PV system operator can quickly bring the underperforming module or modules to optimal performance, thereby, maximizing the power yield of the PV array, and maximizing the revenue of the PV system owner.

ITC/USA 2015 Conference Proceedings / The Fifty-First Annual International Telemetering Conference and Technical Exhibition / October 26-29, 2015 / Bally's Hotel & Convention Center, Las Vegas, NV / In this paper, we develop a miniaturized real time video compression module (VCM) based on embedded motion JPEG 2000 using ADV212 and FPGA. We consider layout of components, values of damping resistors, and lengths of the pattern lines for optimal hardware design. For software design, we consider compression steps to monitor the status of the system and make the system robust. The weight of the developed VCM is approximately 4 times lighter than the previous development. Furthermore, experimental results show that the PSNR is increased about 3dB and the compression processing time is approximately 2 times faster than the previous development.

Modular reinforcement learning is an approach to resolve the curse of dimensionality problem in traditional reinforcement learning. We design and implement a modular reinforcement learning algorithm, which is based on three major components: Markov decision process decomposition, module training, and global action selection. We define and formalize module class and module instance concepts in decomposition step. Under our framework of decomposition, we train each modules efficiently using SARSA($\lambda$) algorithm. Then we design, implement, test, and compare three action selection algorithms based on different heuristics: Module Combination, Module Selection, and Module Voting. For last two algorithms, we propose a method to calculate module weights efficiently, by using standard deviation of Q-values of each module. We show that Module Combination and Module Voting algorithms produce satisfactory performance in our test domain. / text