Thesis (M.S.) University of Missouri-Columbia, 2006. / The entire dissertation/thesis text is included in the research.pdf file; the official abstract appears in the short.pdf file (which also appears in the research.pdf); a non-technical general description, or public abstract, appears in the public.pdf file. Title from title screen of research.pdf file (viewed on August 29, 2007) Includes bibliographical references.

Thesis (MScEng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: The bandwidth of standard HV measurement devices such as capacitive voltage transformers is too limited in order to measure wideband phenomena. This thesis is concerned with the investigation into a non- intrusive HV transient voltage measurement facility using standard substation HV current transformers (CI's) configured in a transconductance topology. The sensing, summation and integration of the CT capacitive earth currents are investigated. This thesis also reports on the development of a optically isolated link using optical fibre for signal transfer and a computer based data acquisition system. / AFRIKAANSE OPSOMMING: Standaard hoogspannings (HS) meettoerusting soos kapasitiewe spannmgs transformators het beperkte bandwydte vir die meet van wyeband verskynsels. Hierdie tesis handel oor die implementering van 'n HS meetstelsel wat op meinbrekende wyse oorgangsverskynsels meet deur middel van HS stroomtransformators wat in 'n transkonduktansie topologie gekonfigureer is. Die meet, sommasie en integrasie van kapasitoewe grondstrome word ondersoek. Hierdie tesis doen ook verslag aangaande die ontwikkeling van 'n optiese geisoleerde koppelvlak wat gebruik word vir seinoordrag en 'n rekenaar gebasseerde data versamelaar. / pdv201406

The Department of Electrical Engineering in fulfilment of the requirements for the Magister Technologiae in Electrical Engineering at the Cape Peninsula University of Technology November 2013 / Pole mounted transformers (PMT) in rural areas present an opportunity for local utilities to do current monitoring on these systems. These transformers are exposed to abnormal amounts of stress due to the vast power demand in these areas. The aim of this study is to develop a more cost-effective condition monitoring system. Transformer current monitoring can be a dangerous practice if not done by suitably trained utility electricians. Hence this study is partly aimed at the elimination of hazardous working environments associated with manual electrical measurements. An investigation to determine a safe and cost-effective way to obtain the electrical measurements required from PMTs is undertaken. Although current measurements can be done with a current clamp-on meter, these measurements still take place at the phases of the transformer and are unsafe. The possibility of implementing wireless data gathering on current clamp-on meters is therefore investigated. This is made possible by a wireless sensor node (WSN) which gathers information and transmits it wirelessly to a WSN base station. This wireless solution is battery powered, necessitating battery replacements, therefore leading to the investigation of magnetic fields, magnetic materials and magnetic induction. A current clamp able to generate a high voltage (HV) output with minimal magnetic field strength is developed. The magnetic fields produced by the transformer’s phase cables are used to generate an alternating voltage. With the help of a microcontroller and an energy harvesting circuit, this voltage is converted and used to charge supercapacitors. The magnetic fields are also used to determine the current flow in the transformer phase cables when the device is not in energy harvesting mode. The device will then undergo comprehensive laboratory testing to determine its accuracy and durability, and is then used to do ‘real life’ current measurements, the results of which are compared against an off-the-shelf current monitoring device.

Orientador: José Pissolato Filho / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-21T11:22:49Z (GMT). No. of bitstreams: 1 Carriel_LazaroPartamian_M.pdf: 4727086 bytes, checksum: dcf745210139d03b3d6ea5af9c6db4d8 (MD5) Previous issue date: 2012 / Resumo: O transformador é o maior ativo, o mais importante e o mais caro do sistema elétrico de potência. Este estudo tem a finalidade de apresentar uma metodologia e um processo de análise do ciclo de vida útil de transformadores de potência baseada na condição, para evitar perda de vida adicional devido ao acréscimo de carga em condição de contingência no sistema. A abordagem foi feita em duas fases e abrangem tanto a metodologia da norma ABNT como a norma do IEEE. Na primeira fase é verificado o comportamento do modelo térmico devido a sobrecargas e, na segunda, a influência das condições do óleo isolante / Abstract: The transformer is the largest, most expensive, most important asset in the electric power system. This study will present a methodology to analyze the useful life of power transformers based on their condition, in order to avoid incremental loss of life caused by overload under contingent operating conditions. The approach has been performed in two phases, following both ABNT and IEEE standards. In the first phase, a thermal model evaluates the effect of overloads on transformers' useful life. In the second phase, the influence of the conditions of the insulating oil is verified / Mestrado / Energia Eletrica / Mestre em Engenharia Elétrica

[pt] Sumarização automática é a tarefa de gerar resumos concisos, corretos e com consistência factual. A tarefa pode ser aplicada a diversos estilos textuais, dentre eles notícias, publicações acadêmicas e avaliações de produtos ou lugares. A presente dissertação aborda a sumarização de múltiplas avaliações. Esse tipo de aplicação se destaca por sua natureza não supervisionada e pela necessidade de lidar com a redundância das informações presentes nas avaliações. Os trabalhos de sumarização automática são avaliados utilizando a métrica ROUGE, que se baseia na comparação de n-gramas entre o texto de referência e o resumo gerado. A falta de dados supervisionados motivou a criação da arquitetura MeanSum, que foi a primeira arquitetura de rede neural baseada em um modelo não supervisionado para essa tarefa. Ela é baseada em auto-encoder e foi estendida por outros trabalhos, porém nenhum deles apresentou os efeitos do uso do mecanismo de atenção e tarefas auxiliares durante o treinamento do modelo. O presente trabalho é dividido em duas etapas. A primeira trata de um experimento no qual extensões à arquitetura do MeanSum foram propostas para acomodar mecanismos de atenção e tarefas auxiliares de classificação de sentimento. Ainda nessa etapa, explora-se o uso de dados sintéticos para adaptar modelos supervisionados a tarefas não supervisionadas. Na segunda etapa, os resultados obtidos anteriormente foram utilizados para realizar um estudo sobre o uso de ajuste fino (fine-tuning) de modelos de linguagem Transformers pré-treinados. A utilização desses modelos mostrou ser uma alternativa promissora para enfrentar a natureza não supervisionada do problema, apresentando um desempenho de + 4 ROUGE quando comparado a trabalhos anteriores. / [en] Automatic summarization is the task of generating concise, correct, and factual summaries. The task can be applied to different textual styles, including news, academic publications, and product or place reviews. This dissertation addresses the summary of multiple evaluations. This type of application stands out for its unsupervised nature and the need to deal with the redundancy of the information present in the reviews. The automatic summarization works are evaluated using the ROUGE metric, which is based on the comparison of n-grans between the reference text and the generated summary. The lack of supervised data motivated the creation of the MeanSum architecture, which was the first neural network architecture based on an unsupervised model for this task. It is based on auto-encoder and has been extended to other works, but none explored the effects of using the attention mechanism and auxiliary tasks during training. The present work is divided into two parts: the first deals with an experiment in which we make extensions to the MeanSum architecture, adding attention mechanisms and auxiliary sentiment classification tasks. In the same experiment, we explore synthetic data to adapt supervised models for unsupervised tasks. In the second part, we used the results previously obtained to carry out a second study on fine-tuning pre-trained Transformer language models. The use of these models showed a promising alternative to the unsupervised nature of the problem, outperforming previous works by + 4 ROUGE.

The goal of this project is to find a suitable Few-Shot Learning (FSL) model that can be used in a fault detection system for use in an industrial setting. A dataset of Printed Circuit Board (PCB) images has been created to train different FSL models. This dataset is meant for evaluating FSL models in the specialized setting of fault detection in PCB manufacturing. FSL is a part of deep learning that has seen a large amount of development recently. Few-shot learning allows neural networks to learn on small datasets. In this thesis, various state-of-the-art FSL algorithms are implemented and tested on the custom PCB dataset. Different backbones are used to establish a benchmark for the tested FSL algorithms on three different datasets. Those datasets are ImageNet, PCB Defects, and the created PCB dataset. Our results show that ProtoNets combined with ResNet12 backbone achieved the highest accuracy in two test scenarios. In those tests, the model combination achieved 87.20%and 92.27% in 1-shot and 5-shot test scenarios, respectively. This thesis presents a Few-Shot Anomaly Detection (FSAD) model based on Vision Transformers (ViT). The model is compared to the state-of-the-art FSAD model DevNet on the MVTec-AD dataset. DevNet and ViT are chosen for comparison because they both approach the problem by dividing images into patches. How the models handle the image patches is however very different. The results indicate that ViT Deviation does not obtain as high AUC-ROC and AUC-PR scores as DevNet. This is because of the use of the very deep ViT architecture in the ViT Deviation model. A shallower transformer-based model is believed to be better suited for FSAD. Improvements for ViT Deviation are suggested for future work. The most notable suggested improvement is the use of the FS-CT architecture as a FSAD model because of the high accuracy it achieves in classification. / Målet med detta projekt är att hitta en lämplig Few-Shot Learning(FSL) modell som kan användas i ett feldetekteringssystem för användning i en industriell miljö. Ett dataset av Printed Circuit Board(PCB) bilder har skapats för att träna olika FSL-modeller. Detta datasetär avsedd för att utvärdera FSL-modeller i det specialiserade områdetfeldetektering vid PCB-tillverkning. FSL är en del av djupinlärningsom har utvecklats mycket den senaste tiden. FSL tillåter neuralanätverk att lära sig på små datamängder.I detta examensarbete implementeras och testas olika state-of-theart FSL algoritmer på det anpassade PCB-datasetet. Olika ryggradsmodeller används för att upprätta ett riktmärke för de testade FSL-algoritmernapå tre olika dataset. Dessa dataset är ImageNet[6], PCB Defects[14]och det skapade PCB-datasetet. Våra resultat visar att ProtoNets ikombination med ResNet12-ryggraden uppnådde den högsta noggrannheten i två testscenarier. I dessa tester uppnådde modellkombinationen 87,20% och 92,27% i testscenarier med 1-shot respektive5-shot.Detta examensarbete presenterar en Few-Shot Anomaly Detectionmodell (FSAD) baserad på Vision Transformers (ViT). Modellen jämförs med FSAD-modellen DevNet på MVTec-AD-datasetet. DevNetoch ViT väljs för jämförelse eftersom de båda angriper problemetgenom att dela upp bilder i mindre lappar. Hur modellerna hanterarlapparna är dock väldigt olika. Resultaten indikerar att ViT-Deviationinte får lika hög AUC-ROC och AUC-PR som DevNet. Detta beror påanvändningen av den mycket djupa ViT-arkitekturen i ViT Deviationmodellen. En grundare ViT-baserad modell tros vara bättre lämpadför FSAD. Förbättringar för ViT-Deviation föreslås för framtida arbete.Den mest anmärkningsvärda föreslagna förbättringen är användningen av FS-CT-arkitekturen som en FSAD-modell på grund av de lovande resultaten den uppnår i klassificering.

The world of power systems is ever changing; ever evolving. One such evolution was the advent of Phasor Measurement Units (PMUs). With the introduction of PMUs in the field, power system monitoring and control changed for the better. Innovative and efficient algorithms that used synchrophasors came to be written. To make these algorithms robust, it became necessary to remove errors that crept into the power system with time and usage. Thus the process of calibration became essential when practical decisions started being made based on PMU measurements. In the context of this thesis ‘calibration’ is the method used to estimate a correction factor which, when multiplied with the respective measurement, negates the effect of any errors that might have crept into them due to the instrument transformers located at the inputs of a PMU or the PMU device itself. Though this thesis mainly deals with the calibration of instrument transformers, work has been done previously for calibrating other components of a power system. A brief description of those methods have been provided along with a history on instrument transformer calibration. Three new methodologies for instrument transformer calibration have been discussed in details in this thesis. The first method describes how only voltage transformers can be calibrated by placing optimal number of good quality voltage measurements at strategic locations in the grid, in presence of ratio errors in the instrument transformers and Gaussian errors in the PMUs. The second method provides a way to calibrate all instrument transformers (both current and voltage) in presence of only one good quality voltage measurement located at the end of a tie-line. This method assumes that all the instrument transformers have ratio errors and the PMUs have quantization errors. The third method attains the same objective as the second one, with the additional constraint that the data obtained from the field may be contaminated. Thus, the third method shows how calibration of all the instrument transformers can be done with data that is intermittent and is therefore, the most practical approach (of the three) for instrument transformer calibration. / Master of Science

Made available in DSpace on 2014-07-29T15:08:18Z (GMT). No. of bitstreams: 1 EFEITOS DE CARGAS NAO-LINEARES.pdf: 1238178 bytes, checksum: 7e5ce254fff10ac951edcf5de73046a3 (MD5) Previous issue date: 2012-05-30 / The intense use of solid state converters that use semiconductor in processing electric power of electrical and electronic equipment brings great benefits such as lower cost, greater efficiency and comfort, but also brings disadvantages, because when connected to the distribution system of electricity produced currents distorted relative to the sinusoidal waveform.. Thus also cause nonsinusoidal voltages that produce additional losses in conductors, transformers and also in its own equipment that generated them. In transformers feeding nonlinear loads occur further losses, which increase the temperature and compromising the insulation life. The factors used and internationally recommended in UL 1561-1994 (K-factor) and IEEE Std C57.110-2008 (Harmonic Loss Factor - FHL) to calculate maximum power three-phase transformers operating in these conditions will be objects of comparison. This work, performed in the laboratory with two three-phase transformers connected in the configuration back-to-back, aims to conduct a survey of actual losses in transformers under linear and nonlinear load and also propose a new scale factor that the three-phase transformers feeding these specific charges. Another aspect to be addressed is planning the design of transformers using these factors. The measurements and data processing are performed at low voltages of the transformers and use transducers voltage and current with high accuracy and LabVIEW programming. / O uso intenso de conversores que utilizam semicondutores no processamento da energia elétrica dentro dos equipamentos elétricos e eletrônicos traz grandes benefícios como menor custo, maior eficiência e mais conforto, contudo traz também desvantagens, pois, quando conectados ao sistema de distribuição de energia elétrica produzem correntes distorcidas em relação à forma de onda senoidal. Desta forma, causam também deformações na forma de onda das tensões. Estes efeitos produzem perdas adicionais em condutores, transformadores e também nos próprios equipamentos que as gerou. Em transformadores, quando do suprimento de cargas não-lineares, ocorrem perdas adicionais, que elevam a temperatura comprometendo a isolação e a sua vida útil. Os fatores usados internacionalmente e recomendados na UL 1561-1994 (Fator K) e na IEEE Std C57.110-2008 (Fator de Perda Harmônica - FHL) para cálculo da potência máxima em transformadores trifásicos operando nestas condições serão objetos de comparação. Este trabalho, realizado em laboratório com dois transformadores trifásicos conectados na configuração back-to-back, tem por objetivo realizar um levantamento real das perdas em transformadores submetidos às cargas lineares e não-lineares e também propor um novo fator que dimensione os transformadores trifásicos alimentando estas cargas específicas. Outro aspecto que será abordado é o planejamento do dimensionamento de transformadores utilizando-se destes fatores. As medições e tratamento de dados são realizados nas tensões inferiores dos transformadores que utilizam transdutores de tensão e de corrente de alta precisão e programação em LabVIEW.

Intrinsic value prediction is a Natural Language Processing (NLP) problem consisting in determining a numerical value contained implicitly and non-trivially in a text. In this project, we introduce the SWORDSMAN model (Sentence and Word-level Oracle for Research Documents by Semantic Multi-scale ANalysis), a deep neural network architecture based on transformers whose goal is to predict the cost of research projects from the analysis of their abstract. SWORDSMAN is built on a hybrid structure based on two branches in order to conduct a multi-scale analysis by combining the strengths of global and local perspectives to extract more relevant information from these texts. The local branch uses Convolution Neural Networks (CNNs) to analyse abstracts at fine-grained word level and bring more nuance to the understanding of the context of occurrence of key terms, while the global branch combines Sentence Transformers and Radial Basis Functions (RBFs) to process these abstracts at a higher level to identify the overall context of the project, while being more focused on the content than the form of the data. The joint use of these models allows SWORDSMAN to have a better capacity to understand complex data by using this analysis at different levels of granularity to present a better estimation accuracy. / Förutsägelse av inneboende värde är ett problem inom Natural Language Processing (NLP) som består i att bestämma ett numeriskt värde som finns implicit och icke-trivialt i en text. I det här projektet introducerar vi SWORDSMAN-modellen (Sentence and Word-level Oracle for Research Documents by Semantic Multi-scale ANalysis), en djup neuronal nätverksarkitektur baserad på transformatorer vars mål är att förutsäga kostnaden för forskningsprojekt utifrån analysen av deras abstrakt. SWORDSMAN bygger på en hybridstruktur baserad på två grenar för att genomföra en analys i flera skalor genom att kombinera styrkorna hos globala och lokala perspektiv för att extrahera mer relevant information från dessa texter. I den lokala grenen används CNN-nätverk (Convolution Neural Networks) för att analysera sammanfattningar på finkornig ordnivå och ge mer nyans till förståelsen av sammanhanget för förekomsten av nyckeltermer, medan den globala grenen kombinerar meningstransformatorer och radiella basfunktioner (RBF) för att bearbeta dessa sammanfattningar på en högre nivå för att identifiera projektets övergripande sammanhang, samtidigt som den är mer inriktad på innehållet än på formen av uppgifterna. Den gemensamma användningen av dessa modeller gör det möjligt för SWORDSMAN att ha en bättre förmåga att förstå komplexa data genom att använda denna analys på olika granularitetsnivåer för att presentera en bättre skattningsnoggrannhet. / La prédiction de valeur intrinsèque est un problème de Traitement Automatique du Langage (TAL) consistant à déterminer une valeur numérique contenue de manière implicite et non triviale dans un texte. Dans ce projet, nous introduisons le modèle SWORDSMAN (Sentence and Word-level Oracle for Research Documents by Semantic Multi-scale ANalysis), une architecture de réseaux de neurones profonde basée sur les transformers dont le but est de prédire le coût de projets de recherche à partir de l’analyse de leur abstract. SWORDSMAN est bâti sur une structure hybride reposant sur deux branches afin de mener une analyse multi-échelles en combinant les forces de perspectives globale et locale pour extraire des informations plus pertinentes de ces textes. La branche locale utilise des réseaux de neurones de convolution (CNN) pour analyser les abstracts à l’échelle des mots et apporter plus de nuance à la compréhension du contexte d’apparition des termes clés, là où la branche globale combine Sentence Transformers et fonctions de base radiale (RBF) pour traiter ces abstracts à un plus haut niveau afin d’identifier le contexte général du projet, tout en étant plus focalisée sur le contenu que la forme des données. L’utilisation conjointe de ces modèles permet à SWORDSMAN de disposer d’une meilleure capacité de compréhension de données complexes en se servant de cette analyse à différents niveaux de granularité pour présenter une meilleure précision d’estimation.

The research programme at the University of Canterbury includes the development and applications of partial core inductors and transformers for high voltage testing of generator insulation. Unlike a conventional full core transformer, a partial core transformer has no limbs and yokes. A partial core transformer is a compromise between a full core and coreless transformer. It is superior to its full core counterpart as far as cost, weight and ease of transportation are concerned. Partial core transformers have a low magnetising reactance and hence draw a high magnetising current. This characteristic makes them a perfect fit in applications where the load is capacitive in nature, such as a.c. power frequency high voltage testing of generator insulation and cable testing etc. The work carried out for this thesis focuses on automatically controlling the amount of reactive power on the supply side of a partial core transformer. The considered design includes a third winding around the existing two windings. A power electronic controller is connected to the third winding, which modifies the VAr absorption characteristics of the magnetically coupled supply winding. Two options are considered to achieve continuous reactive power control in the partial core transformer as explained below. First, a thyristor controlled reactor (TCR) is proposed as the VAr controller. It is modelled using PSCAD/EMTDC software. Simulations reveal the design criteria, overall performance and the limitations of the suggested proposal. The TCR connected tertiary winding takes the capacitive burden of the supply. The model demonstrates the ability of the automatically controlled TCR to provide a continuous variation of reactive power without significant under or over compensation. This feature limits the supply current to its real component only, so the supply provides only the losses of the system. Second, a voltage source converter is considered as the VAr controller. This is modelled in PSCAD/EMTDC and a hardware prototype is designed and built. Based on the analysis, the control algorithm (including a digital PI controller) is implemented using an 8 bit micro-controller, PIC18LF4680. The prototype is tested in the laboratory for both active and inductive load conditions as seen from the supply side. Performance of the hardware prototype is discussed in detail. The PSCAD/EMTDC model and the hardware prototype successfully demonstrate the feasibility of a STATCOM controlled partial core transformer. The proposed system is capable of compensating a wide range of capacitive loads as compared with its TCR counterpart. It is proved that the system is very robust and remains dynamically stable for a large system disturbance such as change in load from full capacitive to inductive and vice versa. This confirms that the system is capable of providing continuous VAr control.