Today's Network Operation Centers (NOC) consist of teams of network professionals responsible for monitoring and taking actions for their network's health. Most of these NOC actions are relatively complex and executed manually; only the simplest tasks can be automated with rules-based software. But today's networks are getting larger and more complex. Therefore, deciding what action to take in the face of non-trivial problems has essentially become an art that depends on collective human intelligence of NOC technicians, specialized support teams organized by technology domains, and vendors' technical support. This model is getting increasingly expensive and inefficient, and the automation of all or at least some NOC tasks is now considered a desirable step towards autonomous and self-healing networks. In this work, we investigate whether such decisions can be taken by Artificial Intelligence instead of collective human intelligence, specifically by Deep-Reinforcement Learning (DRL), which has been shown in computer games to outperform humans. We build an Action Recommendation Engine (ARE) based on RL, train it with expert rules or by letting it explore outcomes by itself, and show that it can learn new and more efficient strategies that outperform expert rules designed by humans by as much as 25%.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / Foi desenvolvido um modelo hidrÃulico, implementado na linguagem de programaÃÃo Delphi, para simulaÃÃo de redes de abastecimento de Ãgua, que utiliza o MÃtodo das CaracterÃsticas (MOC), introduzindo novas capacidades ao modelo CALHIDREDMT de Righetto (1994). O modelo se baseia no conceito de pseudo-transiente, utilizando um valor fictÃcio para a celeridade da onda no MOC. Foram realizadas simulaÃÃes em redes equipadas com reservatÃrios de nÃvel fixo e variÃvel (tanques), estaÃÃes elevatÃrias e boosters, considerando situaÃÃes estÃticas e de perÃodo estendido (24 horas), para casos reais e exemplos genÃricos de redes. Os resultados do modelo desenvolvido foram comparados Ãs simulaÃÃes realizadas no modelo EPANET e com dados de pressÃo medidos em manÃmetro instalado para experimento de campo. Foram avaliados aspectos de convergÃncia do MOC nos elementos de fronteira e o desempenho dessa tÃcnica de soluÃÃo para diferentes configuraÃÃes de rede. AlÃm disso, o modelo hidrÃulico foi acoplado a uma rotina de otimizaÃÃo baseada em Algoritmos GenÃticos (AG). A rotina de otimizaÃÃo apresenta uma funÃÃo objetivo formulada para minimizaÃÃo dos custos energÃticos nas estaÃÃes elevatÃrias, considerando a penalizaÃÃo de aspectos indesejados da operaÃÃo, como pressÃes e velocidades elevadas, excesso de manobras nas bombas e violaÃÃo de limites em tanques. Foram realizadas anÃlises quanto Ãs formas de tarifaÃÃo e variaÃÃo dos parÃmetros do AG, avaliando a influÃncia desses aspectos sobre o valor da funÃÃo objetivo e o tempo de processamento. Foi possÃvel concluir que o modelo apresenta desempenho satisfatÃrio e bastante flexibilidade, tendo potencial para aplicaÃÃes em operaÃÃo de redes pressurizadas de distribuiÃÃo em tempo real. / A hydraulic model was developed for simulation of water supply networks using the Method of Characteristics (MOC), introducing new features to the CALHIDREDMT model, proposed by Righetto (1994). The model uses the concept of pseudo-transient, using a fictitious value for the wave celerity. Simulations were performed in networks equipped with reservoirs, tanks, pump stations and boosters, considering both steady-state situations, as well as extended-periods (24 hours), including real cases and generic examples. The obtained results were compared to simulations performed in the model EPANET and to field experiments (pressure data measured with a manometer). Convergence issues and the performance of the solution technique were analyzed in the boundary elements for different network configurations. In addition, the hydraulic model was coupled into a Genetic Algorithms (GA) routine. The Genetic Algorithm routine presents an objective function that considers the minimization of the energetic costs associated with pump stations and it includes ways to penalize undesired issues in the operation, like high values of pressure and velocity, surplus of pump station maneuvers and violation of tank capacity limits. Analyses were performed considering variations in the taxation formulas, penalization criterions and GA parameters to evaluate the influence of these issues in the objective function and the routine processing time. The model presented a satisfactory performance and flexibility. It can be used as an analytical tool for applications involving water distribution networks in real time.

The increasing awareness on the effect of carbon emissions in our planet has led to several countries to adopt targets for their reduction. One way of contributing to this aim is to use and distribute electricity more efficiently. In this context, Conservation Voltage Reduction (CVR), a well-known technique that takes advantage of the positive correlation between voltage and demand to reduce energy consumption, is gaining renewed interest. This technique saves energy by only reducing customer voltages, without relying on customer actions and, therefore, can be controlled by the Distribution Network Operator (DNO). CVR not only brings benefits to the electricity system by reducing generation requirements (fewer fossil fuel burning and carbon emissions), but also to customers, as energy bill reductions. The extent to which CVR can bring benefits mainly depends on the customers load composition and their voltages. While the former dictates the voltage-demand correlation, the latter constraints the voltage reduction that can be applied without violating statutory limits. Although CVR has been studied for many years, most of the studies neglect the time-varying voltage-demand characteristic of loads and/or do not assess end customer voltages. While these simplifications could be used to estimate CVR benefits for fixed and limited voltage reductions, realistic load and network models are needed to assess the performance of active CVR schemes, where voltages are actively managed to be close to the minimum limit. Moreover, distribution networks have been traditionally designed with limited monitoring and controllability. Therefore, CVR has been typically implemented by adopting conservative voltage reductions from primary substations, for both American and European-style networks. However, as new infrastructure is deployed in European-style LV networks (focus of this work), such as monitoring and on-load tap changers (OLTCs), the opportunity arises to actively manage voltages closer to end customer (unlocking further energy savings). Although these technologies have shown to effectively control voltages in LV networks, their potential for CVR has not been assessed before. Additionally, most CVR studies were performed in a context where distributed generation (DG) was not common. However, this has changed in many countries, with residential photovoltaic (PV) systems becoming popular. As this is likely to continue, the interactions of residential PV and CVR need to be studied. This thesis contributes to address the aforementioned literature gaps by: (i) proposing a simulation framework to characterise the time-varying voltage-demand correlation of individual end customers; (ii) developing a process to model real distribution networks (MV and LV) from DNO data; (iii) adopting a Monte Carlo-based quantification process to cater for the uncertainties related to individual customer demand; (iv) assessing the CVR benefits that can be unlocked with new LV infrastructure and different PV conditions. To accomplish (iv), first, a simple yet effective rule-based scheme is proposed to actively control voltages in OLTC-enabled LV networks without PV and using limited monitoring. It is demonstrated that by controlling voltages closer to customers, annual energy savings can increase significantly, compared to primary substation voltage reductions. Also, to understand the effect of PV on CVR, a centralized, three-phase AC OPF-based CVR scheme is proposed. This control, using monitoring, OLTCs and capacitors across MV and LV networks, actively manages voltages to minimize energy consumption in high PV penetration scenarios whilst considering MV-LV constraints. Results demonstrate that without CVR, PV systems lead to higher energy imports for customers without PV, due to higher voltages. Conversely, the OPF-based CVR scheme can effectively manage voltages throughout the day, minimising energy imports for all customers. Moreover, if OLTCs at secondary substations are available (and managed in coordination with the primary substation OLTC), these tend to regulate customer voltages close to the minimum statutory limit (lower tap positions), while the primary OLTC delivers higher voltages to the MV network to also reduce MV energy losses.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The increasing number of users connected to the Internet led it to become a major vehicle for personal and business transactions in the last years. Nevertheless, its unavailability can result in losses, including nancial ones, for its users. Despite of all eorts to keep the network availability nearest to 100% of the time, reasearches have shown that the existing protocols have two algorithmic problems caused by message losses or disruption, named No Brain and Split Brain, which attack the network availability and lead it to crash. Thus, those researches propose that such protocols must be changed considering the possibility of message loss. In this way, this research species and implements the High Availability Router Protocol (HARP), which is a new high availability protocol that operates in stateless environments. Furthermore, a validation system is presented to test high availability protocols for the sake of link failures. The specication concerns to environment assumptions, services, vocabulary, format and procedure rules specied by nite state machine, moreover, the specication is complemented with a TLA+ formal description regarding concurrent systems context intending to ratify the HARP good properties. The HARP implementation consists of its prototyping on FPGA and the validation system based on a System-on-Programmable Chip (SOPC). / O crescente número de usuários conectados à Internet favoreceu que ela se tornasse um dos principais veículos de transações pessoais e empresariais nos últimos anos. Entretanto, sua indisponibilidade pode acarretar perdas, inclusive de caráter nanceiro, aos seus usuários. Apesar dos esforços empenhados para manter a rede 100% do tempo dispon ível, pesquisas apontam que os protocolos de alta disponibilidade apresentam problemas algorítmicos conhecidos como Acéfalo e Cérebro Partido, que são causados por perdas e erros de mensagens e levam à indisponibilidade da rede. Tais pesquisas propõem, então, que alterações sejam feitas nas especicações dos protocolos existentes considerando que mensagens podem não chegar a seus destinos conforme previsto. Em virtude disso, este trabalho especica e implementa um novo protocolo de alta disponibilidade, chamado High Availability Router Protocol (HARP), cuja operação acontece em ambientes sem preservação de estado. Adicionalmente, apresenta-se um sistema de validação para protocolos de alta disponibilidade que os testam segundo falhas nos canais de comunicação. A especicação do HARP concerne ao ambiente de operação, serviços, vocabulário, formato de mensagens e regras de procedimento especicadas através de máquinas de estados - nitos. Ademais, a especicação é complementada pela descrição formal em TLA+ e sua vericação no contexto de sistemas concorrentes para raticar as boas propriedades do protocolo. A implementação do HARP consiste da prototipagem em FPGA e o sistema de validação é baseado em um System on a Programmable Chip. / Mestre em Ciência da Computação

Made available in DSpace on 2016-04-04T18:31:22Z (GMT). No. of bitstreams: 1 Andrey Guedes Oliveira.pdf: 4354812 bytes, checksum: 17825c1f558c7ff18c82b21e9935d9fb (MD5) Previous issue date: 2008-02-14 / This work aims to describe the measurement the operational indicators in services attending and support problems realized by Network Operation Center, as well as, to establish attendance metrics directed towards contract maintenance, as well as, recommending the use of the simulations in the Operational Plans of the Service Information Technology Providers. The use of the Theory of Lines, eTOM (enhanced Telecom Operations Map) and ITIL (Information Technology Infrastructure Library) assisted in the analysis of the behavior of Network Operation Center. Had been used a real data from a Network Operation Center of the Datacenter, that it pertaining to a Services Telecommunications Provider. However, it is observed that the Theory of Lines and the simulations need a maximum value of attendance to be applied to the historical data of an operational team. The simulation by this attendance parameter allowed us to map the capacity of the Datacenter team with eight analysts, within acceptable values to fulfill Service Level Agreements up to 92% precision to support problems, 85% to analyze alarm tickets and 89% of utilization in a services attendance. The others simulations had projected the behavior of the team with two new scenes: ten and twelve analysts. The use of the computational simulations can be analyzed and be compared the simulated real data and with attendance projections, making possible an operational planning adjusted the new based business-oriented modalities in convergent networks. The computer simulations using maximum parameter, allowed a analyzing of services types behavior and establishing an operational planning. / Este trabalho tem por objetivo dimensionar recursos humanos nos Centros de Opera??es de Redes e estabelecer indicadores de atendimento voltados para a manuten??o dos contratos e a melhoria dos servi?os prestados, bem como utilizar simula??es computacionais no planejamento operacional para o Centro de Opera??es de Redes. As an?lises de atendimentos deste Centro de Opera??es foram realizadas por meio das melhores pr?ticas em gest?o de Tecnologia da Informa??o - ITIL (Information Technology Infrastructure Library), pelo mapa de processos de Telecomunica??es - eTOM (enhanced Telecom Operations Map) e pela Teoria das Filas. Para isto, foram utilizados dados hist?ricos e reais de um Centro de Opera??es de Rede de um Datacenter, pertencente a um Provedor de Servi?os de Tecnologia da Informa??o e Comunica??o. Foram analisados os dados de atendimentos via Teoria das Filas e comparado com os dados simulados. Notouse a necessidade de um valor m?ximo de atendimento para o dimensionamento da equipe, e esse par?metro serviu de base em novas simula??es que possibilitaram uma an?lise comparativa com proje??es de atendimento. A capacidade do grupo com os patamares aceit?veis, ao cumprimento dos Acordos de N?veis de Servi?os com oito analistas, ocorreu dentro de uma precis?o m?dia de 92% para solu??es a problemas, 85% para suporte a alarmes nos equipamentos e 89% de utiliza??o nas presta??es de servi?os. As demais simula??es projetaram o comportamento da equipe com dois novos cen?rios: dez e doze analistas. Os resultados propiciaram o estabelecimento de um planejamento operacional com os principais indicadores de atendimento e informa??es relevantes para a tomada de decis?es empresariais.

Orientador: Marcos Julio Rider Flores / Abstract: Electric distribution networks should be prepared to provide an economic and reliable service to all customers, as well as to integrate technologies related to distributed generation, energy storage, and plug-in electric vehicles. A proper representation of the electric distribution network operation, taking into account smart grid technologies, is key to accomplish these goals. This work presents mathematical formulations for the steady-state operation of electric distribution networks, which consider the unbalance of three-phase grids. Mathematical models of the operation of smart grid-related devices present in electric distribution networks are developed (e.g., volt-var control devices, energy storage systems, and plug-in electric vehicles). Furthermore, features related to the voltage dependency of loads, distributed generation, and voltage and thermal limits are also included. These formulations constitute a mathematical framework for optimization analysis of the electric distribution network operation, which could assist planners in decision-making processes. Different objectives related to technical and/or economic aspects can be pursued within the framework; in addition, the extension to multi-period and multi-scenario optimization is discussed. The presented models are built based on mixed integer linear programming formulations, avoiding the use of conventional mixed integer nonlinear formulations. The application of the presented framework is illustrated throughou... (Complete abstract click electronic access below) / Resumo: As redes de distribuição de energia elétrica devem estar preparadas para fornecer um serviço econômico e confiável a todos os clientes, bem como para integrar tecnologias relacionadas à geração distribuída, armazenamento de energia e veículos elétricos. Uma representação adequada da operação das redes de distribuição, considerando as tecnologias de redes inteligentes, é fundamental para atingir esses objetivos. Este trabalho apresenta formulações matemáticas para a operação em regime permanente das redes de distribuição, que consideram o desequilíbrio de redes trifásicas. Modelos matemáticos da operação de dispositivos relacionados à redes inteligentes presentes em redes de distribuição são desenvolvidos (e.g., dispositivos de controle volt-var, sistemas de armazenamento de energia e veículos elétricos). Além disso, características relacionadas à dependência da tensão das cargas, geração distribuída e limites térmico e de tensão também estão incluídos. Essas formulações constituem um marco matemático para a análise de otimização da operação das redes de distribuição de energia elétrica, o que possibilita modelar os processos de tomada de decisões. Objetivos diferentes relacionados a aspectos técnicos e/ou econômicos podem ser almejados dentro deste marco; Além disso, a extensão para otimização multi-período e multi-cenário é discutida. Os modelos apresentados são construídos com base em formulações de programação linear inteira mista, evitando o uso de formulações não-lineare... (Resumo completo, clicar acesso eletrônico abaixo) / Doutor

The future operation of public low voltage networks has to consider increased decentralised generation using photovoltaic systems for residential application. For utilities it is necessary to determine the maximum permissible installed power of residential photovoltaic systems in sections of the low-voltage network. For this purpose a method based on modelling low-voltage network structures, occurring loads and insolation situations is presented and demonstrated. The maximum permissible installed power of residential photovoltaic systems is fundamentally determined by the structure of the settlement of the affected low-voltage network section. By modifying the generator model the method can be applied too for other types of decentralized generators in the low-voltage network such as fuel cell systems. / Für den Betrieb des öffentlichen Niederspannungsnetzes in Wohnsiedlungsgebieten ist zukünftig mit einem verstärkten Einsatz dezentraler photovoltaischer Energiewandlungsanlagen zu rechnen. Für Netzbetreiber ist es erforderlich, die maximal zulässige Netzanschlussleistung derartiger Anlagen für Niederspannungs-Netzbezirke zu bestimmen. Dazu wird ein Verfahren auf der Grundlage der Modellierung der Struktur von Netzbezirken, der dort auftretenden Belastungen und Einstrahlungssituationen vorgestellt und demonstriert. Die maximal zulässige Netzanschlussleistung dezentraler photovoltaischer Energiewandlungsanlagen wird wesentlich durch die Siedlungsstruktur des betreffenden Niederspannungs-Netzbezirkes bestimmt. Durch Modifikation des Erzeugermodelles kann das Verfahren auch für andere dezentrale Kleinerzeuger im Niederspannungsnetz angewandt werden.

Submitted by CARLOS FRANCISCO SABILLON ANTUNEZ (cfsa27@gmail.com) on 2018-05-23T00:13:45Z No. of bitstreams: 1 20180522ATeseCarlos.pdf: 6005665 bytes, checksum: cc46f4ea50bb15771fa5c7f3ce3f8107 (MD5) / Approved for entry into archive by Cristina Alexandra de Godoy null (cristina@adm.feis.unesp.br) on 2018-05-24T14:51:28Z (GMT) No. of bitstreams: 1 sabillonantunez_cf_dr_ilha.pdf: 6416516 bytes, checksum: 8832a0c7772aecb5081e9daa768de603 (MD5) / Made available in DSpace on 2018-05-24T14:51:28Z (GMT). No. of bitstreams: 1 sabillonantunez_cf_dr_ilha.pdf: 6416516 bytes, checksum: 8832a0c7772aecb5081e9daa768de603 (MD5) Previous issue date: 2018-03-26 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / As redes de distribuição de energia elétrica devem estar preparadas para fornecer um serviço econômico e confiável a todos os clientes, bem como para integrar tecnologias relacionadas à geração distribuída, armazenamento de energia e veículos elétricos. Uma representação adequada da operação das redes de distribuição, considerando as tecnologias de redes inteligentes, é fundamental para atingir esses objetivos. Este trabalho apresenta formulações matemáticas para a operação em regime permanente das redes de distribuição, que consideram o desequilíbrio de redes trifásicas. Modelos matemáticos da operação de dispositivos relacionados à redes inteligentes presentes em redes de distribuição são desenvolvidos (e.g., dispositivos de controle volt-var, sistemas de armazenamento de energia e veículos elétricos). Além disso, características relacionadas à dependência da tensão das cargas, geração distribuída e limites térmico e de tensão também estão incluídos. Essas formulações constituem um marco matemático para a análise de otimização da operação das redes de distribuição de energia elétrica, o que possibilita modelar os processos de tomada de decisões. Objetivos diferentes relacionados a aspectos técnicos e/ou econômicos podem ser almejados dentro deste marco; Além disso, a extensão para otimização multi-período e multi-cenário é discutida. Os modelos apresentados são construídos com base em formulações de programação linear inteira mista, evitando o uso de formulações não-lineares inteiras mistas convencionais. A aplicação do marco apresentado é ilustrada em abordagens de controle para coordenação de carregamento de veículos elétricos, controle de magnitude de tensão e controle de geração distribuída renovável. Diversos métodos são desenvolvidos, com base no marco de otimização matemática, para otimizar a operação de sistemas de distribuição desbalanceados, considerando não apenas diferentes penetrações de veículos elétricos e fontes de energia renováveis, mas também a presença de sistemas de armazenamento e dispositivos de controle volt-var. A este respeito, o agendamento dinâmico e a otimização multi-período de janela rolante são frequentemente usados para alcançar uma operação ótima na rede. A eficácia e robustez das metodologias, bem como a confiabilidade do marco de otimização matemática, são verificados usando vários sistemas de teste (e.g., 123-node, 34-node e 178-node) com nós de média e baixa tensão, diferentes janelas de controle e várias disponibilidades de controle relacionadas aos dispositivos de rede inteligente. / Electric distribution networks should be prepared to provide an economic and reliable service to all customers, as well as to integrate technologies related to distributed generation, energy storage, and plug-in electric vehicles. A proper representation of the electric distribution network operation, taking into account smart grid technologies, is key to accomplish these goals. This work presents mathematical formulations for the steady-state operation of electric distribution networks, which consider the unbalance of three-phase grids. Mathematical models of the operation of smart grid-related devices present in electric distribution networks are developed (e.g., volt-var control devices, energy storage systems, and plug-in electric vehicles). Furthermore, features related to the voltage dependency of loads, distributed generation, and voltage and thermal limits are also included. These formulations constitute a mathematical framework for optimization analysis of the electric distribution network operation, which could assist planners in decision-making processes. Different objectives related to technical and/or economic aspects can be pursued within the framework; in addition, the extension to multi-period and multi-scenario optimization is discussed. The presented models are built based on mixed integer linear programming formulations, avoiding the use of conventional mixed integer nonlinear formulations. The application of the presented framework is illustrated throughout control approaches for plug-in electric vehicle charging coordination, voltage magnitude control, and renewable distributed generation control. Several methods are developed, based on this framework, to optimize the operation of unbalanced distribution systems considering not only different penetrations of electric vehicles and renewable energy sources but also the presence of storage systems and volt-var control devices. In this regard, dynamic scheduling and rolling multi-period optimization are often used to achieve optimal economic operation in the grid. The effective and robustness of the methodologies, as well as the reliability of the mathematical framework, are verified using many test systems (e.g., 123-node, 34-node, and 178-node) with medium and low voltage nodes, different operation control time frames, and several control availabilities related to the smart grid devices.

The future operation of public low voltage networks has to consider increased decentralised generation using photovoltaic systems for residential application. For utilities it is necessary to determine the maximum permissible installed power of residential photovoltaic systems in sections of the low-voltage network. For this purpose a method based on modelling low-voltage network structures, occurring loads and insolation situations is presented and demonstrated. The maximum permissible installed power of residential photovoltaic systems is fundamentally determined by the structure of the settlement of the affected low-voltage network section. By modifying the generator model the method can be applied too for other types of decentralized generators in the low-voltage network such as fuel cell systems. / Für den Betrieb des öffentlichen Niederspannungsnetzes in Wohnsiedlungsgebieten ist zukünftig mit einem verstärkten Einsatz dezentraler photovoltaischer Energiewandlungsanlagen zu rechnen. Für Netzbetreiber ist es erforderlich, die maximal zulässige Netzanschlussleistung derartiger Anlagen für Niederspannungs-Netzbezirke zu bestimmen. Dazu wird ein Verfahren auf der Grundlage der Modellierung der Struktur von Netzbezirken, der dort auftretenden Belastungen und Einstrahlungssituationen vorgestellt und demonstriert. Die maximal zulässige Netzanschlussleistung dezentraler photovoltaischer Energiewandlungsanlagen wird wesentlich durch die Siedlungsstruktur des betreffenden Niederspannungs-Netzbezirkes bestimmt. Durch Modifikation des Erzeugermodelles kann das Verfahren auch für andere dezentrale Kleinerzeuger im Niederspannungsnetz angewandt werden.

This Master's thesis deals with solution of nonstandard operation connection of E.ON's 22kV distribution network in given area, considering the possibilities of operating dispersed electric power sources currently available, as well as possible power sources connected to the distribution network in the future and its potential operation limitation. A part of this paper is a description of calculation methods used for network evaluation in steady state, with emphasis on linear calculations used by E-Vlivy program, in which the operational possibilities are simulated. Therefore, a description of this program is also included. Suggested operation possibilities correspond with E.ON's valid distribution network operation rules. The outcome of this paper will serve to E.ON's 22kV network's controllers as materials for network operating in nonstandard operation states.