In the aviation industry, safety and robustness are the number one priorities, which is why they use well-tested systems such as hydraulic actuators. However, drawbacks such as high weight and maintenance have pushed the industry toward newer, electrical, actuators that are more efficient and lighter. Electrical actuators, on the other hand, have some reliability issues. In particular, short circuits in the stator windings of Permanent-Magnet SynchronousMotors (PMSMs), referred to as Inter-Turn Short Faults (ITSFs), are the dominating faults, and is the focus of this thesis. ITSFs are usually challenging to detect and often do not become noticeable until the fault has propagated, and the motor is on the verge of being destroyed. This thesis investigates the possibility of detecting ITSFs in a PMSM, at an early stage when only one turn is shorted. The method is limited to finding the faults using ML algorithms. Both an experiential PMSM and a simulated model of the experimental PMSM, with the ability to induce an ITSF, were used to collect the data. Several Machine Learning (ML) models were developed, and then trained and tested with the collected data. The results show that four of the tested ML models, being: Random Forest, Gaussian SVM, KNN, and the CNN, all achieve an accuracy exceeding 95%, and that the fault can be found at an early stage in a PMSM with three coils connected in parallel in each phase. The results also show that the ML models are able to identify the ITSF when the simulated data is downsampled to the same frequency as the experimental data. We conclude that the ML models, provided in this study, can be used to detect an ITSF in a simulated PMSM, at an early stage when only one turn is shorted, and that there is great potential for them to detect ITSFs in a physical motor as well.

Les limiteurs de courants de court-circuit supraconducteurs sont des appareillages à fort potentiel pour les réseaux électriques. En effet, ils offrent une limitation efficace dès les premiers instants du court-circuit. On peut qualifier la limitation de "naturelle", c’est-à-dire qu’elle est intrinsèquement liée aux caractéristiques du matériau et ne nécessite pas de commande particulière. Afin de faciliter la conception et l’intégration des limiteurs de courant de court-circuit supraconducteurs résistifs (rSFCL) destinés aux réseaux électriques, il est nécessaire de disposer de modèles de simulation précis. Ces derniers doivent prendre en compte et simuler correctement (et le plus précisément possible) les phénomènes électriques et thermiques du rSFCL en présence de surintensités de courant, qu’il s’agisse d’un court-circuit franc ou d’un phénomène temporaire de plus faible amplitude. Il est difficile d’envisager la planification de l’intégration d’un rSFCL sans passer par des outils numériques qui permettent la simulation d’un tel dispositif dans un réseau électrique en régime transitoire. Il est alors plus facile d’appréhender et de prédire le comportement transitoire du limiteur dans des conditions de stress réalistes, qui peuvent comprendre une grande variété de surintensités, tant en durée qu’en amplitude. Néanmoins, les rSFCL sont des dispositifs fortement non-linéaires caractérisés par un couplage électrique et thermique très fort. L’implémentation d’un tel modèle dans un logiciel de simulation de type “circuits électriques” en régime transitoire présente un certain défi. Bien que des modèles de rSFCL existent déjà, des améliorations doivent être apportées pour prendre en compte i) l’ensemble des phénomènes physiques liés à la limitation (thermiques et électriques), ii) les propriétés géométriques des rubans supraconducteurs utilisés et iii) la possibilité de réaliser des études globales (impact du limiteur sur le réseau) et iv) l’influence de l’architecture du ruban en présence de phénomènes locaux (points chauds). Cette thèse se concentre donc sur le développement d’un modèle de rSFCL basé sur des rubans supraconducteurs de deuxième génération. Ce modèle est développé dans le logiciel EMTP-RV, qui est un outil utilisé par un grand nombre de compagnies d’électricité dans le monde. Le modèle proposé dans cette thèse repose sur une analogie qui fait le lien entre les phénomènes électriques et thermiques, et qui permet une modélisation entièrement basée sur des éléments de circuits électriques. Le modèle permet de prendre également en compte les propriétés non linéaires des matériaux, tant au niveau électrique qu’au niveau thermique, avec l’utilisation de dipôles non-linéaires. Le modèle a été développé pour offrir un niveau de généricité intéressant pour la modélisation des rubans supraconducteurs. Il permet un fonctionnement avec une excitation AC ou DC en tension ou en courant et tient compte de la non-uniformité de courant critique, qui est typiquement observée dans la longueur des rubans disponibles commercialement. Il est également possible de représenter des variantes d’architectures (géométries et matériaux), avec une souplesse de modélisation qui est basée sur un assemblage de blocs “élémentaires” dont les dimensions peuvent être différentes. Cela permet alors d’évaluer, dans une même simulation, l’architecture du limiteur à une échelle submillimétrique (points chauds) et à une échelle “systémique”, tel que le comportement de plusieurs centaines de mètres de ruban. Des comparaisons ont permis de vérifier que le modèle circuit avait un comportement similaire à son équivalent en éléments finis, seulement si la taille des éléments électrothermiques de base (dans EMTP-RV) est adéquate. Le modèle équivalent circuit permet de réaliser des simulations de différentes architectures de rubans supraconducteurs, avec ou sans résistance d’interface, entre les couches tampons et la couche de (RE)BCO par exemple [...] / Superconducting fault current limiters (SFCL) are a promising technology for power systems, i.e. they provide efficient current limitation from the very beginning of the fault without requiring any control system. In fact, the current limiting characteristics are directly connected to the physical properties of superconducting materials. There is a need for accurate models to help designing resistive-type SFCLs (rSFCL) and planning their integration into electrical networks. Such models have to take into account the physics involved for simulating (as accurately as possible) the electrical and thermal behaviours for a wide range of fault conditions, i.e. high and low short-circuit currents that can be of various durations. It is difficult to see how the planning and integration of SFCLs can be realized without using numerical tools, especially tools that allow realizing power system transient simulations, such as EMTP-RV. In fact, such software packages support engineers in predicting the behaviour of SFCLs in realistic network conditions, which may comprise a wide variety of overcurrent or fault situations. However, rSFCLs exhibit highly non-linear behaviours with a strong coupling between thermal and electrical phenomena. The implementation of such a model in power systems simulation tools is therefore challenging. Although some models have been already developed over the years, improvements are needed to take into account i) all the phenomena linked to the current limitation (electrical and thermal), ii) geometric properties of superconducting tapes that are used in rSFCLs, and iii) the possibility to perform simulations at the system level, and iv) the influence of the tape architecture in relationship to local phenomena (hot spots). This thesis hence focuses on the development of a models for resistive-type SFCLs based on second generation high temperature superconducting coated conductors (2G HTS CCs), i.e. (RE)BCO tapes. The models are implemented in EMTP-RV, a tool that is used by many utilities around the world. However, the modeling technique can be adapted to other simulation tools as well. The model proposed in this thesis is based on an electro-thermal analogy, which allows modeling thermal effects with non-linear electrical circuit elements such as resistors and capacitors. The model has been developed with the aim of providing flexibility. Hence, it can be used with an AC or DC excitation, and can also take into account non-uniformity in critical critical current density typically observed along length of the conductors (i.e. tapes). It also allows modeling virtually any tape architecture using modular and flexible electrical and thermal basic building blocks that can be different in size. This in turn also allows modeling SFCLs with different level of discretization, i.e. from hot spot modeling with local heat transfer to several meters of (RE)BCO tape. It therefore becomes possible to analyze in the same simulation phenomena happening at the sub-millimetric scale, such as hot-spot phenomena, and at the system-scale, such as the impact on the network of several hundred meters of superconducting tape. In order to validate the EMTP-RV circuit model, comparisons with results obtained with finite elements have been carried out. A similar behavior could be observed, as long as the discretization size of the electro-thermal elements were appropriate. The EMTP-RV circuit model allows performing optimizations of the tape architecture for various thicknesses of stabilizer, in presence or not of an interfacial resistance layer, e.g. between the superconductor and the substrate. While the circuit model was developed to allow representing heat transfer and current distribution in 3D, simulations are still limited to 2D cases because the size of the nodal matrix is otherwise exceeded in EMTP-RV. Simulation results also show that neglecting heat transfer along the thickness of the tape can be risky, [...]

Neste trabalho estudou-se o movimento de uma carga espacial sobre uma matriz fixa, de densidade uniforme, tanto em circuito aberto como fechado. No primeiro caso, circuito aberto, a solução é quase trivial comparada com o outro caso, no qual o problema é finalmente reduzido a uma equação diferencial ordinária, com método de solução análogo ao empregado em problemas de carga espacial monopolar livre. Como ilustração, estudou-se a voltagem e a corrente produzidas por um sistema com carga total nula, mas com excessos locais de carga / In this work the motion of a space charge cloud embedded in a matrix of Constant immobile charge density is studied in open as well as in closed circuit. In the first case, open circuit, the solution is almost trivial as compared as the other one in which, after some work, the problem is reduced to an ordinary diffetential equation. The method of solution is parallel to that employed in the study of monopolar free space charge motion. The voltage and the current produced by a system with no net charge but with unbalanced local charge density were calculated using the general equations derived in the first part of the work

Malgré d'importants progrès ces dix dernières années, les piles à combustible de type PEMFC (à membrane échangeuse de protons) souffrent toujours de fin de vie prématurée. Le catalyseur et la membrane, principaux constituants du cœur de la pile, sont les deux éléments principalement mis en cause. Ce travail a pour objectif de comprendre les modes de rupture et de dégradation de la membrane électrolyte durant le fonctionnement. Différents types de vieillissement ont été analysés, à la fois en laboratoire mais également sur des systèmes ayant fonctionné sur site en conditions réelles d'opération (jusqu'à 13000 heures). Au travers une approche multi-échelle (du système macroscopique à l'analyse des propriétés macromoléculaires de la membrane) et d'une utilisation systématique (plusieurs centaines d'échantillons analysés), des scénarios de dégradation ont été établis. Dans un premier temps, des outils de caractérisation macroscopiques ont été spécifiquement développés pour sonder rapidement l'ensemble des cellules d'un stack. Ces outils permettent d'identifier les défauts inter et intra-cellule tout en discriminant les propriétés barrières aux gaz des propriétés d'isolation électronique des membranes, tous deux responsables des courants de fuite en système. Cette approche systématique sur l'ensemble des échantillons a mis en évidence des zones spécifiques favorisant la dégradation prématurée des membranes. Dans un second temps, des caractérisations physico-chimiques ciblées dans ces zones de défaillance ont révélé une dégradation fortement localisée et principalement favorisée par des conditions opératoires spécifiques dans les zones d'entrée des gaz. / In spite of strong improvements in fuel cell design this last ten years, Proton Exchange Membrane Fuel Cell are still suffering of premature end of life. Failure of the heart of fuel cell, composed of membrane and catalysts, is commonly responsible for fuel cell shutdown. This work brings an original contribution in understanding membrane degradation mechanisms. Different ageing tests were analyzed, in laboratory as well as in real life operating conditions (up to 13000 hours of solicitations). Within a multi-scale approach, from macroscopic to microscopic, and with a systematic usage (hundreds of samples fully characterized), some degradation mechanisms were established. Firstly, macroscopic tools were specifically developed to rapidly track state of health of all the cells from each stack. With the help of these tools, we were able to identify defects inter and intra-cell. It was also possible to discriminate between gas crossover or electronic short-circuit defects, both responsible for current leaks. This systematic approach on each samples put forward some specific areas within the membrane where degradation was promoted. Secondly, physico-chemical characterizations were performed on membrane targeted areas. It was shown that membrane degradation is strongly localized in some specific channels of the bipolar plates and favored by specific operating conditions in the gaz inlets areas.

A instalação de geração distribuída nas redes de distribuição e de subtransmissão de energia elétrica tem apresentado significativo crescimento em âmbito mundial, impulsionada pelos benefícios que pode proporcionar aos sistemas elétricos, pela necessidade de diversificação da matriz energética dos países, pela desregulamentação do setor de energia elétrica em diversos países e pela necessidade de gerar energia elétrica de forma sustentável. No entanto, para que os geradores distribuídos possam de fato beneficiar a operação das redes elétricas, seus impactos técnicos devem ser cuidadosamente estudados, sobretudo em redes de distribuição, as quais foram inicialmente planejadas para operar com fluxo de potência unidirecional. Dentre esses impactos, destaca-se a elevação do nível de curto-circuito da rede, pois correntes de curto-circuito com valores elevados podem causar violação das capacidades dos equipamentos em suportar os esforços térmicos e dinâmicos e também provocar a perda da coordenação entre os dispositivos de proteção contra sobrecorrente. Neste contexto, este trabalho analisa a influência de geradores distribuídos baseados em inversores na corrente de curto-circuito trifásica em um sistema de distribuição de energia, e sobretudo, comprova que sua contribuição no valor da corrente de falta não supera o dobro de sua corrente nominal. Além disso, com base nesta comprovação, apresenta duas estratégias para inserção desse tipo de gerador nos cálculos de curto-circuito de um sistema de distribuição de energia elétrica. As duas estratégias apresentam resultados satisfatórios, utilizam conceitos básicos de circuitos elétricos, independem de dados minuciosos dos inversores e são validadas através de aplicações numéricas com resultados confrontados aos obtidos das simulações. / The installation of distributed generation in distribution and subtransmission systems has shown significant growth worldwide, driven by the benefits it can provide to electrical systems, the need to diversify the energy sources, deregulation of the electricity industry in several countries and the need to generate electricity in a sustainable manner. However, in order to evaluate if distributed generators benefit the operation of power networks, their technical impacts should be carefully studied, especially in distribution networks, which were originally designed to operate with unidirectional power flow. Among the aforementioned impacts, there is the increase of the short circuit level in the distribution network, since high short-circuit currents may exceed the capabilities of equipment to support the dynamic and thermal stresses and also cause loss of coordination between the overcurrent protection devices. Within this context, this thesis analyzes the influence of inverter based distributed generators on three-phase short circuit currents in a power distribution systems. Especially it confirms that the contribution to the fault current does not exceed twice its rated value. Furthermore, based on this issue, this work presents two strategies for the inclusion of this type of generator in short circuit calculations. Both strategies presented satisfactory results, use basic concepts of electrical circuits, they do not depend on detailed data from the inverters and the results are validated by using simulation results.

A instalação de geração distribuída nas redes de distribuição e de subtransmissão de energia elétrica tem apresentado significativo crescimento em âmbito mundial, impulsionada pelos benefícios que pode proporcionar aos sistemas elétricos, pela necessidade de diversificação da matriz energética dos países, pela desregulamentação do setor de energia elétrica em diversos países e pela necessidade de gerar energia elétrica de forma sustentável. No entanto, para que os geradores distribuídos possam de fato beneficiar a operação das redes elétricas, seus impactos técnicos devem ser cuidadosamente estudados, sobretudo em redes de distribuição, as quais foram inicialmente planejadas para operar com fluxo de potência unidirecional. Dentre esses impactos, destaca-se a elevação do nível de curto-circuito da rede, pois correntes de curto-circuito com valores elevados podem causar violação das capacidades dos equipamentos em suportar os esforços térmicos e dinâmicos e também provocar a perda da coordenação entre os dispositivos de proteção contra sobrecorrente. Neste contexto, este trabalho analisa a influência de geradores distribuídos baseados em inversores na corrente de curto-circuito trifásica em um sistema de distribuição de energia, e sobretudo, comprova que sua contribuição no valor da corrente de falta não supera o dobro de sua corrente nominal. Além disso, com base nesta comprovação, apresenta duas estratégias para inserção desse tipo de gerador nos cálculos de curto-circuito de um sistema de distribuição de energia elétrica. As duas estratégias apresentam resultados satisfatórios, utilizam conceitos básicos de circuitos elétricos, independem de dados minuciosos dos inversores e são validadas através de aplicações numéricas com resultados confrontados aos obtidos das simulações. / The installation of distributed generation in distribution and subtransmission systems has shown significant growth worldwide, driven by the benefits it can provide to electrical systems, the need to diversify the energy sources, deregulation of the electricity industry in several countries and the need to generate electricity in a sustainable manner. However, in order to evaluate if distributed generators benefit the operation of power networks, their technical impacts should be carefully studied, especially in distribution networks, which were originally designed to operate with unidirectional power flow. Among the aforementioned impacts, there is the increase of the short circuit level in the distribution network, since high short-circuit currents may exceed the capabilities of equipment to support the dynamic and thermal stresses and also cause loss of coordination between the overcurrent protection devices. Within this context, this thesis analyzes the influence of inverter based distributed generators on three-phase short circuit currents in a power distribution systems. Especially it confirms that the contribution to the fault current does not exceed twice its rated value. Furthermore, based on this issue, this work presents two strategies for the inclusion of this type of generator in short circuit calculations. Both strategies presented satisfactory results, use basic concepts of electrical circuits, they do not depend on detailed data from the inverters and the results are validated by using simulation results.

Protection of both the wind plant and the interconnecting transmission system during short-circuit faults is imperative for maintaining system structural integrity and reliability. The circuit breakers and protective relays used to protect the power system during such events are designed based upon calculations of the current that will flow in the circuit during the fault. Sequence-network models of various power-system components, such as synchronous generators, transformers, transmission lines, etc., are often used to perform these calculations. However, there are no such models widely accepted for certain types of wind-turbine generators used in modern wind plants. The problem with developing sequence-network models of wind plants is that several different wind-turbine generator designs exist; yet, each exhibit very different short-circuit behavior. Therefore, a “one size fits all” approach is not appropriate for modeling wind plants, as has been the case for conventional power plants based on synchronous-generator technology. Further, many of the newer wind-turbine designs contain proprietary controls that affect the short-circuit behavior, and wind-turbine manufacturers are often not willing to disclose these controls. Thus, protection engineers do not have a standard or other well-established model for calculating short-circuit currents in power systems with wind plants. Therefore, the research described in this dissertation involves the development of such models for calculating short-circuit currents from wind-turbine generators. The focus of this dissertation is on the four existing wind-turbine generator designs (identified as Types 1 – 4). Only AC-transmission-interconnected wind-turbine generators are considered in this dissertation. The primary objective of this research is the development of sequence-network models, which are frequency-domain analysis tools, for each wind-turbine generator design. The time-domain behavior of each wind-turbine generator is thoroughly analyzed through transient simulations, experimental tests on scaled wind-turbine generator test beds, and solutions to the system dynamic equations. These time-domain analyses are used to support the development of the sequence-network models.

Orientador: Jozué Vieira Filho / Banca: Anna Diva Plasencia Lotufo / Banca: Marco Aparecido Queiroz Duarte / Resumo: Este trabalho apresenta uma metodologia para detecção e classificação de faltas do tipo curto-circuito. Esta operação de diagnóstico tem como principal operador um banco de redes neurais artificiais (RNAs) do tipo ARTMAP Nebulosa. Tal ferramenta segue a principal filosofia das RNAs: o reconhecimento de padrões, através de um trei- namento supervisionado inicial, responsável por atualizar os parâmetros das redes con- comitantemente com os resultados desejados. Finalizada a fase de adaptação, as redes serão capazes de, após receber novos padrões, classificá-los, de forma a propiciar, ao operador, informações importantes quanto ao estado de transmissão de barras presentes em um determinado complexo elétrico. Aqui, tomou-se como modelo a simulação de um grande alimentador real, composto por mais de 800 barras. Aliados às RNAs, tam- bém foram aplicadas, no que se refere à análise de dados amostrados de corrente elétri- ca, duas importantes e flexíveis ferramentas matemáticas: a Transformada de Fourier e a Transformada Wavelet. .Observa-se, ao final do trabalho, que os resultados apresenta- dos são bastante encorajadores, o que possibilitaria a utilização do programa em um ambiente real / Abstract: This work presents a detection and classification of short-circuit faults metho- dology. The main operator of this diagnostic operation is a Fuzzy ARTMAP Artificial Neural Networks (ANN) bank. This tool follows the ANN's main core: pattern recogni- tion, through an initial supervised training stage, responsible for updating the parame- ters of the networks con-comitant with the desired results. Completed the adaptation phase, the network will be able to, after receiving new standards, classifying them in order to provide to the operator important information about the state of transmission of bars present in a given complex electric. Here, there is, like a model, a simulation of a real big feeder, comprising more than 800 bars. Besides the ANNs, two important and flexible mathematical tools were also applied with regard to the analysis of sampled data of electric current: Fourier Transform and Wavelet Transform. At the end of the work the results presented are very encouraging, which would allow the use of the pro- gram in a no simulations real environment / Mestre

Submitted by Gustavo Nascimento (gustavo.diniz@ufcg.edu.br) on 2018-02-06T14:28:39Z No. of bitstreams: 1 EDUARDO JORGE CALHEIROS JUNIOR - DISSERTACAO PPGEE 2014.pdf: 3431884 bytes, checksum: 5d6e954510b9e7405666b81ef4f13f30 (MD5) / Made available in DSpace on 2018-02-06T14:28:39Z (GMT). No. of bitstreams: 1 EDUARDO JORGE CALHEIROS JUNIOR - DISSERTACAO PPGEE 2014.pdf: 3431884 bytes, checksum: 5d6e954510b9e7405666b81ef4f13f30 (MD5) Previous issue date: 2014-06 / As análises de adequabilidade de equipamentos são indispensáveis para permitir que as empresas do setor elétrico conheçam as condições em que os mesmos operam, garantindo assim o desempenho dos ativos da empresa e sua disponibilidade no Sistema Interligado Nacional. Dentre os equipamentos analisados, destacam-se os transformadores de potência pelo seu alto custo e a sua importância para o sistema elétrico. Grande parte das falhas em transformadores de potência é de origem dielétrica, relacionadas com a danificação da isolação dos condutores devido às deformações mecânicas ocasionadas pelas altas correntes de curto-circuito passantes nos enrolamentos, reduzindo a vida útil do equipamento. A partir destas considerações, propôs-se avaliar os esforços eletromecânicos internos a alguns transformadores de potência presentes no sistema CHESF, por intermédio de simulações computacionais, quando da ocorrência de curto-circuito trifásico. A metodologia para as simulações consistiu em selecionar os transformadores a serem analisados e coletar as informações técnicas necessárias fornecidas pelos fabricantes. Posteriormente foram determinadas as condições sistêmicas às quais os transformadores estariam submetidos e obtidos os níveis de curto-circuito resultantes, para então efetuar as simulações propriamente ditas. Os resultados das simulações computacionais mostraram os esforços eletromecânicos estimados em quatro transformadores de potência em operação no sistema elétrico da CHESF, resultantes do efeito dinâmico da corrente de curto-circuito trifásico que percorrem os enrolamentos do transformador, considerando diferentes configurações sistêmicas. As conclusões das análises dos resultados indicaram possíveis casos de superações de limites admissíveis e revelou a importância dos aspectos construtivos dos transformadores na sua capacidade de suportar os esforços mecânicos causados por curtos-circuitos. / Analysis of suitability of equipment are indispensable for the electric sector companies meet the conditions in which they operate, thus ensuring the performance of the company's assets and its availability in the National Interconnected System (Sistema Interligado Nacional). The power transformers are included among the analyzed equipment due to its high cost and importance for the electrical system. Most faults in power transformers is of dielectric origin, related to insulation damage due to mechanical deformations, caused by the high short-circuit currents in the windings, reducing the life of the equipment. Then, it was proposed to evaluate the internal electromechanical efforts to some power transformers present in the CHESF system, through computer simulations, in the event of a three-phase short circuit. The methodology for the simulations consisted to select the transformers to parse and collect the necessary technical information provided by the manufacturers. Later, certain systemic conditions in which the transformers would be subjected were defined, it was obtained the resulting short circuit levels and then was performed the simulations themselves. The results of the computer simulations showed the electromechanical efforts estimated at four power transformers in operation in the CHESF electrical system, resulting from the dynamic effect of three-phase short-circuit current that run through the windings of the transformer, considering different systemic configurations. The conclusions of the analysis results indicated possible cases of overruns of permissible limits and reveal the importance of the transformers constructive aspects in their ability to withstand the mechanical stress caused by short circuits.

Made available in DSpace on 2014-06-11T19:22:32Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-02-18Bitstream added on 2014-06-13T18:49:30Z : No. of bitstreams: 1 biscaro_aap_me_ilha.pdf: 1057545 bytes, checksum: feb928c10ad58cb9fc7bef33003e6afc (MD5) / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / Neste trabalho é proposta uma técnica de otimização para alocar medidores dispersos de tensão em alimentadores radiais aéreos de distribuição baseada na metaheurística Greedy Randomized Adaptive Search Procedure (GRASP), com o objetivo de melhorar o desempenho de algoritmos de localização de faltas que utilizam informações esparsas de tensão e, simultaneamente, fazer o controle em tempo real do nível de tensão do alimentador operando sob diferentes cenários. No modelo de programação proposto para alocar os medidores considera-se a natureza estocástica das variáveis envolvidas no problema relacionado com o estudo de contingências em sistemas de energia elétrica, ou seja: carregamento dos transformadores da rede no instante de ocorrência da falta, impedância de falta, probabilidade de falhas e erros de medição dos medidores dispersos de tensão, probabilidade de ocorrer qualquer um dos tipos de faltas possíveis, entre outros. O modelo da função objetivo contempla a localização de faltas com boa precisão para qualquer algoritmo que utiliza medições esparsas de tensão e os menores valores de magnitudes de tensão no alimentador, operando em condições normais ou sob contingências. Apresentam-se resultados de testes com a metodologia proposta para dois alimentadores reais de distribuição de energia elétrica. O primeiro alimentador é de médio porte, tensão nominal de 13,8 kV e possui 134 barras. O segundo alimentador é de grande porte, tensão nominal de 11,4 kV e possui 3287 barras. / This work proposes an optimization technique to allocate voltage measurement devices on radial overhead electric power distribution feeders based on Greedy Randomized Adaptive Search Procedure (GRASP) metaheuristic, aiming at improving the performance of algorithms for fault location using sparse voltage information, while making the real time control of the feeder’s operating voltage under different scenarios. The proposed programming model to allocate the devices considers the nature of the stochastic variables involved in the problem with the study of contingencies in electric power systems: loading of the network transformers at time of occurrence of failure, fault impedance, probability of failures and errors of measurement of dispersed voltage devices, likelihood that any of the possible types of faults occur, among others. The model's objective function includes the faults location procedure with good precision for any algorithm that uses sparse measurements of voltage and the lowest values of the magnitudes of voltage feeder, operating under normal conditions or contingencies. Test results with the proposed methodology applied to simulated data of two real life feeders are presented. The first feeder is medium size, rated voltage of 13.8 kV and has 134 bars. The second feeder is large, rated voltage of 11.4 kV and has 3287 bars.