Dense linear algebra(DLA) is one of the most seven important kernels in high performance computing. The introduction of new machines from vendors provides us opportunities to optimize DLA libraries for the new machines and thus exploit their power. Unfortunately the optimization phase is not straightforward. The optimum code of a certain Basic Linear Algebra Subprogram (BLAS) kernel, which is the core of DLA algorithms, in two different machines with different semiconductor process can be different even if they share the same features in terms of instruction set architecture, memory hierarchy and clock speed. It has become a tradition to optimize BLAS for new machines. Vendors maintain highly optimized BLAS libraries targeting their CPUs. Unfortunately the existing BLAS for GPUs is not highly optimized for DLA algorithms. In my research, I have provided new algorithms for several important BLAS kernels for different generation of GPUs and introduced a pointer redirecting approach to make BLAS run faster in generic problem size. I have also presented an auto-tuning approach to parameterize the developed BLAS algorithms and select the best set of parameters for a given card. The hardware trends have also brought up the need for updates on existing legacy DLA software packages, such as the sequential LAPACK. To take advantage of the new computational environment, successors of LAPACK must incorporate algorithms of three main characteristics: high parallelism, reduced communication, and heterogeneity-awareness. On multicore architectures, Parallel Linear Algebra Software for Multicore Architectures (PLASMA) has been developed to meet the challenges in multicore. On the other extreme, Matrix Algebra on GPU and Multicore Architectures (MAGMA) library demonstrated a hybridization approach that indeed streamlined the development of high performance DLA for multicores with GPU accelerators. The performance of these two libraries depend upon right choice of parameters for a given problem size and given number of cores and/or GPUs. In this work, the issue of automatically tuning these two libraries is presented. A prune based empirical auto-tuning method has been proposed for tuning PLASMA. Part of the tuning method for PLASMA was considered to tune hybrid MAGMA library.

Dense linear algebra(DLA) is one of the most seven important kernels inhigh performance computing. The introduction of new machines from vendorsprovides us opportunities to optimize DLA libraries for the new machinesand thus exploit their power. Unfortunately the optimization phase is notstraightforward. The optimum code of a certain Basic Linear AlgebraSubprogram (BLAS) kernel, which is the core of DLA algorithms, in twodifferent machines with different semiconductor process can be differenteven if they share the same features in terms of instruction setarchitecture, memory hierarchy and clock speed. It has become a traditionto optimize BLAS for new machines. Vendors maintain highly optimized BLASlibraries targeting their CPUs. Unfortunately the existing BLAS for GPUsis not highly optimized for DLA algorithms. In my research, I haveprovided new algorithms for several important BLAS kernels for differentgeneration of GPUs and introduced a pointer redirecting approach to makeBLAS run faster in generic problem size. I have also presented anauto-tuning approach to parameterize the developed BLAS algorithms andselect the best set of parameters for a given card.The hardware trends have also brought up the need for updates on existinglegacy DLA software packages, such as the sequential LAPACK. To takeadvantage of the new computational environment, successors of LAPACK mustincorporate algorithms of three main characteristics: high parallelism,reduced communication, and heterogeneity-awareness. On multicorearchitectures, Parallel Linear Algebra Software for MulticoreArchitectures (PLASMA) has been developed to meet the challenges inmulticore. On the other extreme, Matrix Algebra on GPU and MulticoreArchitectures (MAGMA) library demonstrated a hybridization approach thatindeed streamlined the development of high performance DLA for multicoreswith GPU accelerators. The performance of these two libraries depend uponright choice of parameters for a given problem size and given number ofcores and/or GPUs. In this work, the issue of automatically tuning thesetwo libraries is presented. A prune based empirical auto-tuning method hasbeen proposed for tuning PLASMA. Part of the tuning method for PLASMA wasconsidered to tune hybrid MAGMA library.

The modern day mechanical integration and organisation of cities have disturbed the natural systems and sequences of the environment, as well as the social fabric and collective notions of society within cities (Mumford, 1938:17). Man's progressive "liberation" from the natural environment has desensitised him to nature (Crowe, 1995:233) creating an artifi cial realm where man is removed from nature. Pretoria's establishment was greatly infl uenced by its landscape - aptly called "the valley between the ridges and the rivers" (Dippenaar, 2013:6). Urban modernisation and infrastructure development have however progressively disregarded the river and the landscape within the city, consequently detaching the city from these natural elements, which resulted in the loss of their signifi cance. As one of the last remaining underdeveloped sites in the inner city of Pretoria, Berea Park represents a continuation of the landscape and the Apies River channel into the city. Situated at the southern gateway of the city, Berea Park was Pretoria's fi rst offi cial recreational, social and cultural venue, hosting numerous sport, outdoor and other events. Unfortunately Berea Park is currently in a derelict and idle state. Infrastructural developments have gradually disconnected the site from the surrounding city fabric and community. The isolated and forgotten park, has like the river, tragically lost its essential signifi cance. The divide between the city fabric, the Apies River channel and the Berea Park landscape presents the opportunity to explore architecture's role in re-integrating and connecting experiences of "nature" into the city to enhance the quality of urban life. This dissertation explores architecture and the hybrid typology of "building as a landscape", as a means to create a new topography and interface between the Berea Park landscape, river channel and the city. The concept of a hybrid architecture that merges the typologies of urban park, landscape, infrastructure, public amenity and building is explored with the intention to develop a new typology that can respond to the contextual challenges. The building becomes an extension of the park, and the park becomes an extension of the building. The new interface, proposed programme and activities are aimed at linking the city, its inhabitants, communities, the landscape and the river. / Die hedendaagse meganiese integrasie en organisasie van stede het die natuurlike sisteme en ritmes van die omgewing, asook die sosiale strukture en kollektiewe opvattings van gemeenskappe in stede versteur (Mumford, 1038:17). Die mens se progressiewe "bevryding" van die natuurlike omgewing het gelei tot 'n skeiding tussen die mens en die natuur (Crowe, 2013:6), wat der volglik ? kunsmatige domein waarbinne die mensdom uit die natuurverwyder is, geskep het. Die totstandkoming van Pretoria was grootliks be?nvloed deur sy natuurlike landskap - met die gepaste benaming van " 'n vallei gele? tussen rante en riviere" (Dippenaar, 2013:6). Met stedelike modernisering en infrastruktuur ontwikkeling is daar stelselmatig minder ag geslaan op die rivier en die omliggende stadslandskap, wat daartoe gelei het dat die stad van die natuurlike omgewing afgesonder geraak het en uiteindelik het dit die verlies van die essensi?le betekenis en waarde van die natuurelemente veroorsaak. Berea Park, een van die laaste oorblywende onderontwikkelde terreine in die sentrale stadsgebied van Pretoria, is gele? aan die suidelike ingang van die stad en verteenwoordig 'n verlenging van die landskap en die Apiesrivierkanaal na die stad. Di? terrein was Pretoria se eerste amptelike ontspannings-, sosiale en kulturele byeenkomsplek, wat sport-, buitelugen ander byeenkomste aangebied het. Ongelukking is Bera Park tans 'n verlate, verwaarloosde en onbenutte area. Infrastruktuur ontwikkeling van di? gebied het geleidelik die terrein afgesonder van die omliggende stad en gemeenskap. Die ge?soleerde en vergete park het, soos die rivier, tragies geen betekenis of nut meer nie. Die skeiding tussen die stedelike bouomgewing, die Apiesrivierkanaal en die landskap van Berea Park bied die geleentheid om argitektuur se rol met betrekking tot die herintegrering en verbinding van die "natuur" met die stad te ondersoek derhalwe die gehalte van 'n stedelike lewenswyse te verhoog. Die skripsie ondersoek argitektuur en die hibriede tipologie van 'n gebou as 'n landskap, wat deur die fasilitering van 'n nuwe topografi e 'n interaktiewe verbinding tussen die landskap van Berea Park, die rivierkanaal en die stad tot stand kan bring. Hibriede-argitektuur is 'n konsep wat die samesmelting van die tipologie? van stedelike park, landskap, infrastruktuur, openbare fasiliteit en gebou ondersoek met die doel om 'n nuwe tipologie te ontwikkel wat kontekstuele uitdagings kan aanspreek. Die gebou word 'n uitbreiding van die park en die park word 'n uitbreiding van die gebou. Die doelwit van die gebou as nuwe interaktiewe verbinding en die voorgestelde program en aktiwiteite, is om die stad se inwoners, gemeenskappe, die landskap en die rivier te herverbind. / Mini Dissertation (MArch (Prof))--University of Pretoria, 2016. / Architecture / MArch (Prof) / Unrestricted

Nous vivons aujourd'hui dans un monde de plus en plus complexe et interconnecté où de nombreuses entités, de plus en plus intelligentes, génèrent une multitude d'interactions qui peuvent contribuer à enrichir leurs capacités. Nous nous intéressons à la collaboration qui favorisera la réalisation de tâches complexes par ces machines d'aujourd'hui et de demain en stimulant ces interactions structurées et en intégrant des processus de décisions intelligents. Par ce biais, cela contribuera à améliorer leur fonctionnement et pourra participer à leur perfectionnement (meilleures connaissances de leur environnement, rapidité d'action et de décision, apport de nouvelles compétences). L'objectif principal de la thèse est donc de contribuer à la compréhension de ce qu'est la collaboration, de sa définition à sa mise en œuvre, en mettant en lumière ses concepts sous-jacents. Nous proposons une méthode d'analyse du besoin et des contraintes ainsi qu'un modèle d'architecture d'agent collaboratif (HACCA) permettant d'intégrer toutes les caractéristiques des processus collaboratifs que nous présentons. Nous présentons également une première implémentation dans la plateforme multi-agents GAMA.Dans le cadre de cette étude, nous nous intéressons à deux cas d'application d'entités mobiles : les drones et les véhicules connectés. Ainsi nous contribuons également plus largement aux besoins d'autonomie et de décision des drones, des véhicules connectés et autonomes du futur, dans un contexte temporel contraint où la qualité des interactions est primordiale pour optimiser le processus de réalisation des objectifs. / We live today in an increasingly complex and interconnected world where many entities, increasingly intelligent, generate a multitude of interactions that can contribute to enrich their capabilities.We are interested in collaboration that will enable complex tasks to be performed by these machines of today and tomorrow by stimulating these structured interactions and integrating intelligent decision-making processes. In this way, it will contribute to improve their functioning and will be able to participate in their improvement (better knowledge of their environment, speed of action and decision-making, provision of new skills).The main objective of the thesis is therefore to contribute to the understanding of what collaboration is, from its definition to its implementation, by highlighting its underlying concepts. We propose a method of analysis (needs and constraints) and a collaborative agent architecture model (HACCA) to integrate all the characteristics of the collaborative processes that we present. We are also showing a first implementation in the GAMA multi-agent platform.As part of this study, we are interested in two cases of application of mobile entities: drones and connected vehicles.Thus we also contribute more to the autonomy needs and decision-making process of drones, connected and autonomous vehicles of the future, in a constrained temporal context where the quality of interactions is essential to optimize the process of achieving objectives.

Optical fibers have brought fast and reliable data transmission to todays network. The immense fiber build-out over the last few years has generated a wide array of new access technologies, transport and network protocols, and next-generation services in the Local Area Network (LAN), Metropolitan Area Network (MAN), and Wide Area Network (WAN). All these different technologies, protocols, and services were introduced to address particular telecommunication needs. To remain competitive in the market, the service providers must offer most of these services, while maintaining their own profitability. However, offering a large variety of equipment, protocols, and services posses a big challenge for service carriers because it requires a huge investment in different technology platforms, lots of training of staff, and the management of all these networks.<p> In todays network, service providers use SONET (Synchronous Optical NETwork) as a basic TDM (Time Division Multiplexing) transport network. SONET was primarily designed to carry voice traffic from telephone networks. However, with the explosion of traffic in the Internet, the same SONET based TDM network is optimized to support increasing demand for packet based Internet network services (data, voice, video, teleconference etc.) at access networks and LANs. Therefore the service providers need to support their Internet Protocol (IP) infrastructure as well as in the legacy telephony infrastructure. Supporting both TDM and packet services in the present condition needs multilayer operations which is complex, expensive, and difficult to manage. A hybrid switch is a novel architecture that combines packets (IP) and TDM switching in a unified access platform and provides seamless integration of access networks and LANs with MAN/WAN networks. The ability to fully integrate these two capabilities in a single chassis will allow service providers to deploy a more cost effective and flexible architecture that can support a variety of different services.<p> This thesis develops a hybrid switch which is capable of offering bundled services for TDM switching and packet routing. This is done by dividing the switchs bandwidth into VT1.5 (Virtual Tributary -1.5) channels and providing SONET based signaling for routing the data and controlling the switchs resources. The switch is a TDM based architecture which allows each switchs port to be independently configured for any mixture of packet and TDM traffic, including 100% packet and 100% TDM. This switch allows service providers to simplify their edge networks by consolidating the number of separate boxes needed to provide fast and reliable access. This switch also reduces the number of network management systems needed, and decreases the resources needed to install, provision and maintain the network because of its ability to collapse two network layers into one platform.<p> The scope of this thesis includes system architecture, logic implementation, and verification testing, and performance evaluation of the hybrid switch. The architecture consists of ingress/egress ports, an arbiter and a crossbar. Data from ingress ports is carried to the egress ports via VT1.5 channels which are switched at the cross point of the crossbar. The crossbar setup and channel assignments at ingress port are done by the arbiter. The design was tested by simulation and the hardware cost was estimated. The performance results showed that the switch is non-blocking, provide differentiated service, and has an overall effective throughput of 80%. This result is a significant step towards the goal of building a switch that can support multiprotocol and provide different network capabilities into one platform. The long-term goal of this project is to develop a prototype of the hybrid switch with broadband capability.

Optical fibers have brought fast and reliable data transmission to todays network. The immense fiber build-out over the last few years has generated a wide array of new access technologies, transport and network protocols, and next-generation services in the Local Area Network (LAN), Metropolitan Area Network (MAN), and Wide Area Network (WAN). All these different technologies, protocols, and services were introduced to address particular telecommunication needs. To remain competitive in the market, the service providers must offer most of these services, while maintaining their own profitability. However, offering a large variety of equipment, protocols, and services posses a big challenge for service carriers because it requires a huge investment in different technology platforms, lots of training of staff, and the management of all these networks.<p> In todays network, service providers use SONET (Synchronous Optical NETwork) as a basic TDM (Time Division Multiplexing) transport network. SONET was primarily designed to carry voice traffic from telephone networks. However, with the explosion of traffic in the Internet, the same SONET based TDM network is optimized to support increasing demand for packet based Internet network services (data, voice, video, teleconference etc.) at access networks and LANs. Therefore the service providers need to support their Internet Protocol (IP) infrastructure as well as in the legacy telephony infrastructure. Supporting both TDM and packet services in the present condition needs multilayer operations which is complex, expensive, and difficult to manage. A hybrid switch is a novel architecture that combines packets (IP) and TDM switching in a unified access platform and provides seamless integration of access networks and LANs with MAN/WAN networks. The ability to fully integrate these two capabilities in a single chassis will allow service providers to deploy a more cost effective and flexible architecture that can support a variety of different services.<p> This thesis develops a hybrid switch which is capable of offering bundled services for TDM switching and packet routing. This is done by dividing the switchs bandwidth into VT1.5 (Virtual Tributary -1.5) channels and providing SONET based signaling for routing the data and controlling the switchs resources. The switch is a TDM based architecture which allows each switchs port to be independently configured for any mixture of packet and TDM traffic, including 100% packet and 100% TDM. This switch allows service providers to simplify their edge networks by consolidating the number of separate boxes needed to provide fast and reliable access. This switch also reduces the number of network management systems needed, and decreases the resources needed to install, provision and maintain the network because of its ability to collapse two network layers into one platform.<p> The scope of this thesis includes system architecture, logic implementation, and verification testing, and performance evaluation of the hybrid switch. The architecture consists of ingress/egress ports, an arbiter and a crossbar. Data from ingress ports is carried to the egress ports via VT1.5 channels which are switched at the cross point of the crossbar. The crossbar setup and channel assignments at ingress port are done by the arbiter. The design was tested by simulation and the hardware cost was estimated. The performance results showed that the switch is non-blocking, provide differentiated service, and has an overall effective throughput of 80%. This result is a significant step towards the goal of building a switch that can support multiprotocol and provide different network capabilities into one platform. The long-term goal of this project is to develop a prototype of the hybrid switch with broadband capability.

Research in mobile robotics, unmanned systems, and autonomous man-portable vehicles has grown rapidly over the last decade. This push has taken the problems of robot cognition and behavioral control out of the lab and into the field. Two good examples of this are the DARPA Urban Challenge autonomous vehicle race and the RoboCup robot soccer competition. In these challenges, a mobile robot must be capable of completing complex, sophisticated tasks in a dynamic, partially observable and unpredictable environment. Such conditions necessitate a behavioral programming approach capable of performing high-level action selection in the presence of multiple goals of dynamically changing importance, and noisy, incomplete perception data. In this thesis, an approach to behavioral programming is presented that provides the designer with an intuitive method for building contextual intelligence while preserving the qualities of emergent behavior present in traditional behavior-based programming. This is done by using a modified hierarchical state machine for behavior arbitration in sequence with a command fusion mechanism for cooperative and competitive control. The presented approach is analyzed with respect to portability across platforms, missions, and functional requirements. Specifically, two landmark case-studies, the DARPA Urban Challenge and the International RoboCup Competition are examined. / Master of Science

A energia elétrica constitui um dos propulsores da economia de um país, assumindo um papel extremamente importante e estratégico, pois influi diretamente na capacidade produtiva. A expansão da produção de energia elétrica não se consegue somente com medidas de curto prazo, pois as obras deste setor demandam um longo tempo de execução, medido em anos e dependendo da magnitude da obra, o prazo pode até superar uma década. O parque gerador nacional é constituído predominantemente por usinas hidroelétricas, complementado por usinas térmicas que utilizam diversas fontes de combustível, havendo a necessidade de minimizar a produção das térmicas, em virtude do alto custo de geração, em relação ao custo de geração hidroelétrica. Garantir o suprimento da demanda futura de energia elétrica é uma tarefa complexa de planejamento que basicamente, depende da análise de dois cenários que se compõem: o primeiro cenário é o que desenha o crescimento futuro da economia e neste caso, desde que não ocorram fatos extraordinários como o recente crescimento econômico experimentado pela China, a previsão da demanda não acarreta surpresa de grande significância; o segundo cenário, traz como característica a incerteza, uma vez que a produção das hidroelétricas depende da quantidade de água disponível dos cursos de água, que por sua vez, depende do regime de chuvas passado e corrente. O índice pluviométrico é um dado estocástico, ocorrendo ao sabor da probabilidade, o que remete a um estudo de casos e seus desdobramentos, acarretando um leque de possibilidades de estados muito grande, dificultando as análises sobre a previsão futura. Planejar o setor elétrico compreende prever um crescimento de demanda e equipar o setor com máquinas de geração, necessárias para atender a demanda, com uma margem de risco calculada. Para isto, utilizam-se modelos de simulação que possibilitem o exercício de previsão, combinando-se os dois cenários citados, visualizando os estados sub seqüentes, decorrentes de decisões tomadas. A dificuldade desta tarefa é devida à quantidade de alternativas da situação futura, resultante de um fenômeno combinatório de possibilidades que exige para simulação dos modelos, não só uma grande capacidade de processamento dos computadores como também, uma estratégia de tratamento do problema, baseada em processos numéricos especializados e dirigidos a este tipo de problema. Dada a importância e magnitude deste assunto, qualquer esforço que venha a contribuir para uma melhoria do planejamento do setor elétrico, traz benefícios significativos, o que corrobora com os propósitos desta tese, que busca em primeiro lugar, propor soluções técnicas viáveis e econômicas para o problema de otimização da geração de energia elétrica, e em segundo lugar, apresentar uma solução para este tipo de problema, com uma abordagem inovadora, provida de um potencial significativo para aplicação em muitos outros tipos de problemas similares. / Electrical power could be considered as one of the economy propulsion vector of a country, assuming extremely important and strategic role because it makes direct influence to the production capacity. The expansion of electrical energy production could not only be done in a short time because constructions in this area take many years and could require more then a decade depending on the magnitude of them. The national power generation group is constituted mainly by hydro power plants complemented by thermal power plants which use several kinds of fuel which generation cost is high, if compared to hydro power generation, and should be minimized. It is a complex planning issue to supply the future power demand which basically depends on the analysis of two compoundable scenarios: the first one refers to the forecast of future economy growing and in this case, unless unpredicted issues occur such as the recent high economy growing experimented by China, the future demand does not show any surprise and is easy to predict; the second one, has inside the uncertainty because the hydro plants productions depends on the water quantity of rivers which depends on the past and current rainfall regimen. The quantity of rainfall is a stochastic data and follows the rules of probability and this drives to the study of cases and its deployments which are numerous causing difficulties to forecast the future. The planning of the electrical area has to examine the future demand and provide the necessary power generation equipment assuming a certain risk. To have it done, simulation models are used to predict the future, combining the two scenarios cited before, and viewing the results promoted by decision took in a step before. The difficult of this task is caused by the big amount of future alternatives provided by the combinatorial phenomena which require, to process the model, a computer with high processing capacity and specialized and specific methods that can resolve this king of problem. Because of the importance and magnitude of this issue, every effort which contributes to the improvement of power planning is welcome and this corroborates with this thesis which has an objective to propose technical, viable and economic solutions to solve the optimization problems with a new approach and has potential to be applied in many others kind of similar problems.

A energia elétrica constitui um dos propulsores da economia de um país, assumindo um papel extremamente importante e estratégico, pois influi diretamente na capacidade produtiva. A expansão da produção de energia elétrica não se consegue somente com medidas de curto prazo, pois as obras deste setor demandam um longo tempo de execução, medido em anos e dependendo da magnitude da obra, o prazo pode até superar uma década. O parque gerador nacional é constituído predominantemente por usinas hidroelétricas, complementado por usinas térmicas que utilizam diversas fontes de combustível, havendo a necessidade de minimizar a produção das térmicas, em virtude do alto custo de geração, em relação ao custo de geração hidroelétrica. Garantir o suprimento da demanda futura de energia elétrica é uma tarefa complexa de planejamento que basicamente, depende da análise de dois cenários que se compõem: o primeiro cenário é o que desenha o crescimento futuro da economia e neste caso, desde que não ocorram fatos extraordinários como o recente crescimento econômico experimentado pela China, a previsão da demanda não acarreta surpresa de grande significância; o segundo cenário, traz como característica a incerteza, uma vez que a produção das hidroelétricas depende da quantidade de água disponível dos cursos de água, que por sua vez, depende do regime de chuvas passado e corrente. O índice pluviométrico é um dado estocástico, ocorrendo ao sabor da probabilidade, o que remete a um estudo de casos e seus desdobramentos, acarretando um leque de possibilidades de estados muito grande, dificultando as análises sobre a previsão futura. Planejar o setor elétrico compreende prever um crescimento de demanda e equipar o setor com máquinas de geração, necessárias para atender a demanda, com uma margem de risco calculada. Para isto, utilizam-se modelos de simulação que possibilitem o exercício de previsão, combinando-se os dois cenários citados, visualizando os estados sub seqüentes, decorrentes de decisões tomadas. A dificuldade desta tarefa é devida à quantidade de alternativas da situação futura, resultante de um fenômeno combinatório de possibilidades que exige para simulação dos modelos, não só uma grande capacidade de processamento dos computadores como também, uma estratégia de tratamento do problema, baseada em processos numéricos especializados e dirigidos a este tipo de problema. Dada a importância e magnitude deste assunto, qualquer esforço que venha a contribuir para uma melhoria do planejamento do setor elétrico, traz benefícios significativos, o que corrobora com os propósitos desta tese, que busca em primeiro lugar, propor soluções técnicas viáveis e econômicas para o problema de otimização da geração de energia elétrica, e em segundo lugar, apresentar uma solução para este tipo de problema, com uma abordagem inovadora, provida de um potencial significativo para aplicação em muitos outros tipos de problemas similares. / Electrical power could be considered as one of the economy propulsion vector of a country, assuming extremely important and strategic role because it makes direct influence to the production capacity. The expansion of electrical energy production could not only be done in a short time because constructions in this area take many years and could require more then a decade depending on the magnitude of them. The national power generation group is constituted mainly by hydro power plants complemented by thermal power plants which use several kinds of fuel which generation cost is high, if compared to hydro power generation, and should be minimized. It is a complex planning issue to supply the future power demand which basically depends on the analysis of two compoundable scenarios: the first one refers to the forecast of future economy growing and in this case, unless unpredicted issues occur such as the recent high economy growing experimented by China, the future demand does not show any surprise and is easy to predict; the second one, has inside the uncertainty because the hydro plants productions depends on the water quantity of rivers which depends on the past and current rainfall regimen. The quantity of rainfall is a stochastic data and follows the rules of probability and this drives to the study of cases and its deployments which are numerous causing difficulties to forecast the future. The planning of the electrical area has to examine the future demand and provide the necessary power generation equipment assuming a certain risk. To have it done, simulation models are used to predict the future, combining the two scenarios cited before, and viewing the results promoted by decision took in a step before. The difficult of this task is caused by the big amount of future alternatives provided by the combinatorial phenomena which require, to process the model, a computer with high processing capacity and specialized and specific methods that can resolve this king of problem. Because of the importance and magnitude of this issue, every effort which contributes to the improvement of power planning is welcome and this corroborates with this thesis which has an objective to propose technical, viable and economic solutions to solve the optimization problems with a new approach and has potential to be applied in many others kind of similar problems.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This work presents a study and analysis of photovoltaic systems subject to partial shading, where the main consequence is the reduction of its energy production. To mitigate this problem, it is proposed a new photovoltaic architecture known as hybrid architecture. In this architecture, the system is separated into arrays that can be partially or completely shaded with other arrays that are not shaded. The partially shaded modules are connected in parallel to maximize its energy production, while unshaded ones are connected in series. A dc-dc converter is used with the array with the modules subject to partial shading for maximum power extraction of this array, to match the operating voltage of both arrays, and to ensure that during its operation, the unshaded array operates in maximum power point. Additionally, it is presented the small signal modeling of the dc-dc converter through the generalized modeling technique, with the inclusion of the dynamic of a photovoltaic array through an equivalent approximate model by a voltage source with a series resistance, and linearized for different curve points of the photovoltaic module. The proposed control system for this converter uses a controller based on the internal model principle, as the resonant one, whose objective is, besides eliminating the low frequency ripple at the input of photovoltaic module to reduce the losses of the maximum power point tracking algorithm, also reduce the dc-bus capacitance in dual-stage photovoltaic inverters. Shaded photovoltaic systems present higher complexity to predict its energy production. This is because there are different configurations to connect photovoltaic modules among each other, which results in different values of generated power. Consequently, a general methodology of energy production forecast that uses only the shaded area of the modules as a factor of analysis can be imprecise. Based on this, a new methodology for energy production forecast in partially shaded photovoltaic systems is presented. The proposed method consists in identifying the shaded modules in each time, to quantify the hourly irradiation incident on all modules, analyze the electrical configuration of the array and, from that, calculate the produced energy. At the end are presented simulation and experimental results that validate the use of hybrid architecture in photovoltaic systems that present partial shading conditions. / Este trabalho apresenta um estudo e análise de sistemas fotovoltaicos que apresentam redução da produção de energia devido ao sombreamento parcial em módulos fotovoltaicos. Para mitigar este problema, é proposta uma nova arquitetura de sistemas fotovoltaicos denominada arquitetura híbrida. Nessa arquitetura, o sistema é separado em um arranjo de módulos que podem ser parcialmente ou totalmente sombreados e outro com módulos não-sombreados. Os módulos do arranjo passíveis de sombreamento são conectados em paralelo entre si para se maximizar sua produção de energia, enquanto que os módulos não-sombreados são arranjados em série. Um conversor CC-CC é empregado junto ao arranjo dos módulos passíveis de sombreamento para extração da máxima potência desse arranjo, para compatibilizar a tensão de operação dos dois arranjos, e para garantir que durante a operação conjunta, o arranjo que nunca é sombreado opere no ponto de máxima potência. Adicionalmente, obtém-se a modelagem de pequenos sinais do conversor CC-CC através da modelagem generalizada, com a inclusão da dinâmica de um arranjo fotovoltaico através de um modelo equivalente aproximado por uma fonte de tensão em série com uma resistência, e linearizado para diversos pontos da curva do painel. Como proposta de sistema de controle desse conversor, utiliza-se um controlador baseado no princípio do modelo interno, como o ressonante, cujo objetivo é, além de eliminar a ondulação em baixa frequência presente na entrada dos painéis para reduzir as perdas do algoritmo de rastreamento do ponto de máxima potência, também possibilitar a redução da capacitância do barramento CC de inversores de duplo estágio. Sistemas fotovoltaicos sombreados apresentam maior complexidade para se prever a produção de energia. Isso ocorre pois existem diversas configurações para se conectar os módulos fotovoltaicos entre si, cujo impacto resulta em diferentes valores da energia gerada. Por consequência, uma metodologia generalizada de previsão da produção de energia que utilize apenas a área sombreada dos módulos como fator de análise pode ser pouco precisa. Com base nisso, uma nova metodologia para previsão da produção de energia em sistemas fotovoltaicos parcialmente sombreados é apresentada. O método proposto consiste em identificar os módulos sombreados em cada hora, quantificar a irradiação horária incidente em todos os módulos, analisar a configuração elétrica do arranjo e, a partir disso, calcular a energia produzida. Por fim são apresentados resultados de simulação e experimentais que validam a utilização da arquitetura híbrida em sistemas fotovoltaicos que apresentam condições de sombreamento parcial.