Water plays a critical role in every aspect of civilization: agriculture, industry, economy, environment, recreation, transportation, culture, and health. Much of America's drinking water and wastewater infrastructure; however, is old and deteriorating. A crisis looms as demands on these systems increase. The costs associated with renewal of these aging systems are staggering. There is a critical disconnect between the methodological remedies for infrastructure renewal problems and the current sequential or isolated manner of renewal analysis and execution. This points to the need for a holistic systems perspective to address the renewal problem. Therefore, new tools are needed to provide support for wastewater infrastructure decisions. Such decisions are necessary to sustain economic growth, environmental quality, and improved societal benefits. Accurate prediction of wastewater pipe structural and functional deterioration plays an essential role in asset management and capital improvement planning. The key to implementing an asset management strategy is a comprehensive understanding of asset condition, performance, and risk profile. The primary objective of this research is therefore to develop protocols and methods for evaluating the wastewater pipe performance. This research presents the life cycle of wastewater pipeline identifying the causes of pipe failure in different phases including design, manufacture, construction, operation and maintenance, and repair/rehabilitation/replacement. Various modes and mechanisms of pipe failure in wastewater pipes were identified for different pipe material which completed with results from extensive literature reviews, and interviews with utilities and pipe associations. After reviewing all relevant reports and utility databases, a set of standard pipe parameter list (data structure) and a pipe data collection methodology were developed. These parameters includes physical/structural, operational/functional, environmental and other parameters, for not only the pipe, but also the entire pipe system. This research presents a development of a performance index for wastewater pipes. The performance index evaluates each parameter and combines them mathematically through a weighted summation and a fuzzy inference system that reflects the importance of the various factors. The performance index were evaluated based on artificial data and field data to ensure that the index could be implemented to real scenarios. Developing a performance index led to the development of a probabilistic performance prediction model for wastewater pipes. A framework would enable effective and systematic wastewater pipe performance evaluation and prediction in asset management programs. / Ph. D.

Performance prediction modeling is a crucial step in assessing the remaining service life of pipelines. Sound infrastructure deterioration models are essential for accurately predicting future performance that, in turn, are critical tools for efficient maintenance, repair and rehabilitation decision making. The objective of this research is to develop a gravity and force main pipe performance deterioration model for predicting the remaining economic life of wastewater pipe for infrastructure asset management. For condition assessment of gravity pipes, the defect indices currently in practice, use CCTV inspection and a defect coding scale to assess the internal condition of the wastewater pipes. Unfortunately, in practice, the distress indices are unable to capture all the deterioration mechanisms and distresses on pipes to provide a comprehensive and accurate evaluation of the pipe performance. Force main pipes present a particular challenge in performance prediction modeling. The consequence of failure can be higher for the force mains relative to the gravity pipes which increases the risk associated with these assets. However, unlike gravity pipes, there are no industry standards for inspection and condition assessment for force mains. Furthermore, accessibility issues for inspections add to this challenge. Under Water Environmental and Reuse Foundation (WEandRF)'s Strategic Asset Management (SAM) Challenge, there was a planned three-phase development of this performance prediction model. Only Phases 1 and 2 were completed for gravity pipes under the SAM Challenge. Currently, 37 utilities nationally distributed have provided data and support for this research. Data standards are developed to capture the physical, operational, structural, environmental, financial, and other factors affecting the performance. These data standards were reviewed by various participating utilities and service providers for completeness and accuracy. The performance of the gravity and force main pipes are assessed with incorporating the single and combined effects of these parameters on performance. These indices assess the performance regarding; integrity, corrosion, surface wear, joint, lining, blockage, IandI, root intrusion, and capacity. These performance indices are used for the long-term prediction of performance. However, due to limitations in historical performance data, an advanced integrated method for probabilistic performance modeling to construct workable transition probabilities for predicting long-term performance has been developed. A selection process within this method chooses a suitable prediction model for a given situation in terms of available historical data. Prediction models using time and state-dependent data were developed for this prediction model for reliable long-term performance prediction. Reliability of performance assessments and long-term predictions are tested with the developed verification and validation (VeandVa) framework. VeandVa framework incorporates piloting the performance index and prediction models with artificial, field, and forensic data collected from participating utilities. The deterioration model and the supporting data was integrated with the PIPEiD (Pipeline Infrastructure Database) for effective dissemination and outreach. / PHD

Volvo trucks are involved in a project with vehicle combinations longer than 25.25 meters and/or with a weight of more than 64 tons. Those combinations are called high capacity transport (HCT). Some of the vehicle combinations are now driven with exemptions for the coupling equipment. The exemption is needed because according to a standard the theoretical transmitted forces between the dolly and the truck exceed the value that the coupling components have been certified for. The theoretical transmitted forces are called performance indexes. This limits the development but the exception could be avoided by using a hinged drawbar on the dolly.  This thesis aims to investigate what needs to be done in order to introduce dolly with hinged drawbar in longer and heavier vehicle combinations. A literature study and interviews were conducted to find problems that could be expected when using hinged drawbar on dolly and to investigate how the performance indexes for a specific vehicle combination should be calculated. Thereafter a dolly was put together through a product development process. Then concepts to solve a problem that was noticed through the literature study were developed. The concepts were sketched, evaluated in a matrix and thereafter the winning concept was modelled in detail in computer aided design (CAD).   According to the traffic definition in Sweden a dolly with hinged drawbar is violating the law. It defines a dolly as a module with rigid drawbar. On the other hand, the European Union defining a dolly as a trailer with fifth- wheel made to convert a semi-trailer into a full trailer. They don’t mention anything that contradicts to the usage of hinged drawbar. So the Swedish definition of a dolly needs to be challenged.   Simulations have shown that the braking performance will be worse when using a dolly with hinged drawbar compared to using a dolly with rigid drawbar. Therefore concepts to improve the braking performance of the dolly have been developed. The concept was to look the fifthwheel from pivoting around a horizontal axle perpendicular to the driving direction. By doing so the dolly and semi-trailer will become more like a full trailer and that would better the braking performance.  Through interviews and literature study it was noticed that the high performance indexes were calculated in an incorrect way. The highest forces are generated through interaction forces which are only affected by the vehicles closest to the coupling components. Therefore a calculation model need to be modified for this. Three calculation models were tested on the vehicle-combination. Two of these generated performance indexes that were reasonable. The calculation models were ISO-combination 3 and a model developed by Sweatman (1980). These two calculation models are modified for vehicle combinations that includes three vehicles. / Volvo lastvagnar ingår i ett forskningsprojekt med lastbilskombinationer som är längre än 25,25  meter och/eller väger mer än 64 ton. Ett sådant fordon kallas högkapacitetstransport (HCT). Idag körs vissa av dessa fordon med dispenser för kopplingsutrustningen. Detta då de teoretiska överförda krafterna mellan lastbil och dolly enligt en standard får värden som är högre än vad kopplingsutrustningen är certifierad med. De teoretiska överförda krafterna benämns som prestandaindex. Denna dispens skulle kunna undvikas genom att använda en ledad dragstång på dollyn.  Detta examensarbete syftar till att undersöka vad som behöver göras för att introducera dolly med ledad dragstång i längre och tyngre fordonskombinationer. En litteraturundersökning samt intervjuer genomfördes för att upptäcka problem som kan uppstå vid användande av dolly med ledad dragstång. Även en undersökning gällande hur prestandaindex ska beräknas på rätt sätt genomfördes.  Därefter konstruerades en dolly genom en produktutvecklingsprocess. Koncept för att lösa ett problem som noterades under litteraturundersökningen utvecklades. Koncepten skissades och blev sedan utvärderade genom en matris. Det vinnande konceptet konstruerades i CAD.   Från litteraturstudien visade det sig att det finns en motstridighet i den svenska definitionen gällande användandet av ledad dragstång på dolly. I Sverige definieras dolly som en kärra. En kärra definieras som en modul med stel dragstång. Däremot definierar EU dolly som en trailer som konverterar en semi-trailer till en full trailer. Där nämns ingenting som kan stoppa användandet av dolly med ledad dragstång. Därför borde den svenska definitionen utmanas.   En dolly med ledad dragstång förväntas ha sämre bromsprestanda än en vanlig dolly. Koncept utvecklades för att förbättra bromsprestandan på dollyn. Konceptet som vann var att låsa vändskivan från att vridas runt en horisontell axel vinkelrätt till åkriktningen. På så viss blir kombinationen dolly, semi-trailer mer som en full trailer vilket skulle innebära en bättre bromsprestanda.  Från intervjuer med kunniga personer och från litteraturstudien framkom det att de höga prestanda indexen var felaktigt beräknade. Det blev även känt att de största krafterna genereras genom interaktionskrafter som endast påverkas av de närmaste fordonens vikt. Tre olika beräkningsmodeller testades där två utav dem gav värden som kan anses vara rimliga. Beräkningsmodellerna var ISO-kombination 3 och en modell utvecklad av Sweatman (1980). Dessa två är anpassad för kombinationer som innehåller 3 fordon.

abstract: After a power system blackout, system restoration is the most important task for the operators. Most power systems rely on an off&ndashline; restoration plan and the experience of operators to select scenarios for the black start path. Using an off&ndashline; designed restoration plan based on past experience may not be the most reliable approach under changing network configurations and loading levels. Hence, an objective restoration path selection procedure, including the option to check constraints, may be more responsive in providing directed guidance to the operators to identify the optimal transmission path to deliver power to other power plants or to pick up load as needed. After the system is subjected to a blackout, parallel restoration is an efficient way to speed up the restoration process. For a large scale power system, this system sectionalizing problem is quite complicated when considering black&ndashstart; constraints, generation/load balance constraints and voltage constraints. This dissertation presents an ordered binary decision diagram (OBDD) &ndashbased; system sectionalizing method, by which the splitting points can be quickly found. The simulation results on the IEEE 39 and 118&ndashbus; system show that the method can successfully split the system into subsystems satisfying black&ndashstart; constraints, generation/load balance constraints and voltage constraints. A power transfer distribution factor (PTDF)&ndashbased; approach will be described in this dissertation to check constraints while restoring the system. Two types of restoration performance indices are utilized considering all possible restoration paths, which are then ranked according to their expected performance characteristics as reflected by the restoration performance index. PTDFs and weighting factors are used to determine the ordered list of restoration paths, which can enable the load to be picked up by lightly loaded lines or relieve stress on heavily loaded lines. A transmission path agent can then be formulated by performing the automatic path selection under different system operating conditions. The proposed restoration strategy is tested on the IEEE&ndash39; bus system and on the Western region of the Entergy system. The testing results reveal that the proposed strategy can be used in real time. / Dissertation/Thesis / Ph.D. Electrical Engineering 2010

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Control systems monitoring relates to ability to supervise the industrial plants operation, evaluating the controller performance by oscillation detection, disturbances, faults in sensors, valve stiction and diagnosing possible causes of problems that may degrade the productive capacity, managing alarms and providing possible strategies to act or even improve the eciency of operation. Control Process Monitoring (CPM) and Control Process Assessment (CPA) technologies are essential for maintaining the ecient during operation of industrial processes. This work presents a systematic method for monitoring of control systems in chemical industry, using a performance manager. This way it was investigated control systems using the performance index concept to quantify the degradation of controller's eciency, that is a tool to evaluate the behavior in industrial operation context. Performance index supplies the idea of as the control system is distant of best control idealized with same structure. Due the characteristics of reference behavior, that does not depend on the structure of control used, it can be evaluated controllers of several types and under several situations. This index was obtained by design of reference, using the factorization with interactor matrix generalized. Discover abnormalities in control systems is very important and complicated. In addition, was evaluated procedures subject to presence of oscillations, using an extension of the Karhunem-Loeve method and periodogramas. It was also evaluated, processes with valves stiction using the Yamashita method and faults in sensors and systems subjected to disturbances, using the principal components analysis (PCA) with T2 and Q statistics, with objective of detecting, isolating and diagnosing possible cause of control systems deterioration. / Monitorar sistemas de controle relaciona-se à capacidade de supervisionar de maneira ampla a operação de plantas industriais, avaliando o desempenho de controladores, detectando oscilações, perturbações, falhas em sensores, emperramento de válvulas de controle e diagnosticando as possíveis causas de problemas que possam degradar a capacidade produtiva, gerenciando alarmes e fornecendo possíveis estratégias de como atuar para manter, ou até mesmo, aprimorar a eficiência da operação. O monitoramento contínuo do desempenho de sistemas de controle (Control Process Monitoring - CPM) e a avaliação do desempenho de sistemas de controle (Control Process Assessment - CPA) são tecnologias essenciais para a manutenção da eficiência durante a operação de processos industriais. Neste contexto, este trabalho apresenta uma sistemática para o monitoramento de sistemas de controle na indústria química, através de um gerenciador de desempenho proposto. Desta forma foi investigado sistemas de controle utilizando o conceito de índice de desempenho para quantificar a degradação da eficiência do controlador, tratando-se de uma ferramenta para a sua avaliação durante a operação industrial. O índice de desempenho fornece a ideia do quanto o controle corrente do sistema esta distante do melhor controle idealizado com mesma estrutura. Devido as características do comportamento de referência, que independe da estrutura de controle utilizada, pode-se avaliar controladores de diversos tipos e sob varias situações. Este índice foi obtido através do projeto da referência, utilizando fatoração de modelos lineares através da matriz de interação generalizada. Descobrir anormalidades constitui-se uma tarefa muito importante e complicada. Desta forma, avaliou-se processos sujeitos à presença de oscilações, utilizando uma extensão do método de Karhunem-Loéve e periodogramas. Avaliou-se também, processos com emperramento de válvulas de controle através do método de Yamashita e falhas em sensores e sistemas submetidos à perturbações, utilizando a analise dos componentes principais (PCA) e as estatísticas T2 e Q, com o objetivo de detectar, isolar e diagnosticar a possível causa da degradação dos sistemas de controle analisados. / Doutor em Engenharia Química

Orientador: Carlos Alberto de Castro Junior / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-07T06:07:11Z (GMT). No. of bitstreams: 1 Dester_Mauricio_M.pdf: 1510731 bytes, checksum: 2d78f28f0e947ab2f8b733433312aa84 (MD5) Previous issue date: 2006 / Resumo: O objetivo deste trabalho é propor um método que permita, de forma rápida e segura, estabelecer uma lista de contingências ordenada por severidade considerando a óptica da estabilidade de tensão. Em linhas gerais, o procedimento consiste em obter o estado da rede após a simulação de uma determinada contingência e calcular o índice de performance correspondente à mesma. Estes passos devem ser repetidos para todos os ramos da rede. Ao final tem-se um índice associado a cada contingência. Para a obtenção da lista ordenada por severidade, classificam-se as contingências considerando como fator de ordenação o valor do índice de performance calculado para cada uma delas. De posse desta lista, pode-se lançar mão de ferramentas que permitam uma análise mais detalhada para o grupo contingências com maior severidade e a tomada de ações no sentido de minimizar as conseqüências causadas por estas contingências. Este método pode ser utilizado como suporte à operação em tempo real, pois permite a seleção das contingências mais severas, de forma rápida, para então posteriormente realizar uma análise mais detalhada das mesmas. Não há, portanto, a necessidade da análise detalhada de todas as contingências da rede, o que permite uma economia de tempo, esforço humano e computacional, recursos estes de fundamental importância considerando a operação em tempo real / Abstract: The aim of this work is to propose a fast and reliable method to rank a list of contingencies ordered by severity under the voltage stability point of view. The basic steps of the method consist of computing the system's post-contingency state and computing the respective performance index. These steps must be repeted for contingencies associated to all system branches. So an ordered contingency list is built considering the performance index of each contingency as a merit factor. With this list it is possible to pick the most severe contingencies and submit them to conventional tools that allow a more complete analysis and appropriate decisions to avoid operating system in insecure conditions if one of the severe contingencies occurs. It is possible to use this method as a real time tool due to its efficiency. Little computational effort is made to get the ordered contingency list and it is possible to focus the human and computational effort on the most severe contingencies. The real time operation requires efficient decision-making procedures and this method brings up the possibility of saving time since it is no longer necessary to analize all system contingencies to find out the most severe one. It suffices to analize the top-ranked contingencies captured by the proposed method / Mestrado / Energia Eletrica / Mestre em Engenharia Elétrica

Made available in DSpace on 2014-08-20T13:59:11Z (GMT). No. of bitstreams: 1 dissertacao_anelise_tessari_perboni.pdf: 3085637 bytes, checksum: c55ca1817bac9993117c80d2a3e5179a (MD5) Previous issue date: 2011-02-14 / Oilseed rape (Brassica napus L.) belongs to the family Brassicaceae and constitutes the third most produced oilseed in the world, surpassed only by soybean and palm. Flooding of soil and low temperatures can cause alterations in cellular metabolism and impair plant growth, generating a stress condition. Two experiments were conducted using the canola hybrid Hyola 43, Hyola 401, Hyola 420 and Hyola 432. The first was conducted in greenhouse where 10 plants of each hybrid were subjected to stress by flooding for a period of six days. Were evaluated parameters and differences in the kinetics of emission of chlorophyll fluorescence, chlorophyll content, leaf area and dry weight of shoot and root. The flooding caused a reduction of leaf area, root dry mass and root dry mass/shoot dry mass ratio of all hybrids studied. In relation to chlorophyll fluorescence were identified differences in parameters related to specific flows that have undergone changes as a function of stress the latter being more pronounced for Hyola 432. The performace photosynthetic index (PIABS and PIABS,total) showed a decrease after 6 days of the experiment except for hybrid Hyola 420. The analysis of the kinetics of chlorophyll a fluorescence transient showed differential effects of stress application in different sites of photosynthetic apparatus, and these effects were less pronounced in the plants of hybrid Hyola 420. In the second experiment, canola plants were grown in a greenhouse where stress by low temperatures was driven in the Laboratory of Plant Metabolism. Ten plants of each hybrid were placed in a growth chamber the absence of light for different periods of low temperatures: 1 and 4 hours at 0oC and 1, 4 and 15 hours at 4oC. Were evaluated parameters and differences in the kinetics of emission of chlorophyll a fluorescence. The hybrids showed the same behavior in relation to chlorophyll a fluorescence in response to stress by low temperatures. In general, the parameters of chlorophyll fluorescence related to the activity of the PSI were the most affected, and the increase of these parameters resulted in higher values of the PIABS,total. The analysis of the kinetics of chlorophyll a fluorescence transient showed temperature effects, mainly at the J point which reflects the accumulation of reduced plastoquinone (QA -) and the I-P phase reflecting the reduction of end electron acceptors at the PSI acceptor side, ie, ferredoxin (Fd) NADP and other intermediaries. We conclude that the plants of the hibrid Hyola 420 were less affected by flooding showing that this hybrid can be considered less sensitive to stress compared to other hybrids. The parameters that involve the activity of the FSI and the variable fluorescence transient at the J point were more responsive to stress conditions imposed by low temperatures. / A canola (Brassica napus L.) pertence à família das Brassicaceae e constitui-se na terceira oleaginosa mais produzida no mundo, superada apenas pela soja e pela palma. O alagamento do solo e as baixas temperaturas podem promover alterações no metabolismo celular e prejudicar o crescimento das plantas, gerando uma condição de estresse. Foram desenvolvidos dois experimentos utilizando os híbridos de canola Hyola 43, Hyola 401, Hyola 420 e Hyola 432. O primeiro foi conduzido em casa de vegetação, onde 10 plantas de cada híbrido foram submetidas a estresse por alagamento por um período de seis dias. Foram avaliados os parâmetros e as diferenças na cinética de emissão da fluorescência da clorofila, teor de clorofila, área foliar e massa seca da parte aérea e raiz. O alagamento promoveu redução da área foliar, massa seca de raízes e relação massa seca de raiz/massa seca parte aérea de todos híbridos estudados. Em relação a fluorescência da clorofila foram identificadas diferenças nos parâmetros relacionados aos fluxos específicos que sofreram variação em função do estresse sendo esta mais acentuada para Hyola 432. Os índices de performace fotossintéticos (PIABS e PIABS,total) apresentaram decréscimo ao final de 6 dias de experimento exceto para o híbrido Hyola 420. A análise da cinética da fluorescência transiente da clorofila a mostrou efeitos diferenciais da aplicação do estresse nos diferentes locais da maquinaria fotossintética, sendo estes efeitos menos pronunciados nas plantas do híbrido Hyola 420. No segundo experimento as plantas de canola foram cultivadas em casa de vegetação sendo o estresse por baixas temperaturas conduzido no Laboratório de Metabolismo Vegetal. Dez plantas de cada híbrido foram acondicionadas em câmara de crescimento com ausência de luz por períodos distintos de baixas temperaturas: 1 e 4 horas a 0oC e 1, 4 e 15 horas a 4oC. Avaliaram-se os parâmetros e as diferenças na cinética de emissão da fluorescência da clorofila a. Os híbridos apresentaram o mesmo comportamento em relação à fluorescência da clorofila em resposta ao estresse por baixas temperaturas. Em geral, os parâmetros da fluorescência da clorofila relacionados à atividade do fotossistema I foram os mais afetados, sendo que o aumento destes parâmetros resultou em maiores valores de PIABS,total. A análise da cinética da fluorescência transiente da clorofila a revelou efeitos da temperatura principalmente no ponto J o qual reflete o acúmulo de plastoquinona A reduzida (QA -) e na fase I-P que reflete a redução dos aceptores de elétrons finais do lado aceptor do FSI, ou seja, ferredoxina (Fd), outros intermediários e NADP. Conclui-se que as plantas do híbrido Hyola 420 foram menos afetadas pelas condições de alagamento demonstrando que tal híbrido pode ser considerado menos sensível a determinado estresse em relação aos demais híbridos testados. Os parâmetros que envolvem a atividade do FSI bem como a 5 fluorescência variável transiente no ponto J foram mais responsivos às condições impostas pelo estresse por baixas temperaturas.

One feature that is hoped for in the smart grid is the participation of energy prosumers in a power market through demand response program. In this work, we consider a third-party virtual power plant (VPP) that has “real-time” control over a number of prosumers’ storage units within an envisaged free market. Typically, a VPP with domestic energy storage will involve a bidirectional flow of energy, where energy can either flow from the grid to the prosumers’ battery or from the prosumers’ battery to the grid. Such a system requires prices to be set correctly in order to meet the market objectives of all the VPP stakeholders (VPP Aggregator, prosumers, and grid). Previous work has shown how VPPs could operate, and the benefits of using energy storage, coupled with pricing, in terms of reducing energy cost for stakeholders and providing the grid with its required load shape. The published work either assumes prices or costs or then optimises for least cost within the grid parameters i.e. losses, voltage limits, etc. However, the setting of prices in such a way that energy can be traded among VPP stakeholders that satisfies all stakeholders’ objectives has not been fully explored in the literature, particularly with real-time VPP aggregators. In this thesis, we present novel strategies for evaluating and setting the prices of a community VPP with domestic storage based on the bidirectional flow of energy through the VPP aggregator between the grid and the prosumers that mutually meet all VPP stakeholders’ objectives. This showed that depending on pricing and the VPP objectives, demand-side management could be attractive. However, the effect on the grid in terms of the load was not what was desired. A new performance index called the “Cumulative Performance Index” CPI is proposed to measure the VPP’s performance. Using the CPI, it was possible to compare and contrast between the VPP technical performance and its business case for stakeholders. Optimizing with respect to the grid’s requirement for DSM from the VPP, it was possible to achieve a CPI of 100%. This work was implemented using a novel approach on a genetic algorithm platform. / Niger Delta Development Commission of Nigeria

Esta tese aborda o problema de obtenção de um índice de carga ou de desempenho adequado para utilização no escalonamento de processos em sistemas computacionais heterogêneos paralelos/distribuídos. Uma ampla revisão bibliográfica com a correspondente análise crítica é apresentada. Essa revisão é a base para a comparação das métricas existentes para a avaliação do grau de heterogeneidade/homogeneidade dos sistemas computacionais. Uma nova métrica é proposta neste trabalho, removendo as restrições identificadas no estudo comparativo realizado. Resultados de aplicações dessa nova métrica são apresentados e discutidos. Esta tese propõe também o conceito de heterogeneidade/homogeneidade temporal que pode ser utilizado para futuros aprimoramentos de políticas de escalonamento empregadas em plataformas computacionais heterogêneas paralelas/distribuídas. Um novo índice de desempenho (Vector for Index of Performance - VIP), generalizando o conceito de índice de carga, é proposto com base em uma métrica Euclidiana. Esse novo índice é aplicado na implementação de uma política de escalonamento e amplamente testado através de modelagem e simulação. Os resultados obtidos são apresentados e analisados estatisticamente. É demonstrado que o novo índice leva a bons resultados de modo geral e é apresentado um mapeamento mostrando as vantagens e desvantagens de sua adoção quando comparado às métricas tradicionais. / This thesis approaches the problem of evaluating an adequate load index or a performance index, for using in process scheduling in heterogeneous parallel/distributed computing systems. A wide literature review with the corresponding critical analysis is presented. This review is the base for the comparison of the existing metrics for the evaluation of the computing systems homogeneity/heterogeneity degree. A new metric is proposed in this work, removing the restrictions identified during the comparative study realized. Results from the application of the new metric are presented and discussed. This thesis also proposes the concept of temporal heterogeneity/homogeneity that can be used for future improvements in scheduling polices for parallel/distributed heterogeneous computing platforms. A new performance index (Vector for Index of Performance - VIP), generalizing the concept of load index, is proposed based on an Euclidean metric. This new index is applied to the implementation of a scheduling police and widely tested through modeling and simulation. The results obtained are presented and statistically analyzed. It is shown that the new index reaches good results in general and it is also presented a mapping showing the advantages and disadvantages of its adoption when compared with the traditional metrics.

Esta tese aborda o problema de obtenção de um índice de carga ou de desempenho adequado para utilização no escalonamento de processos em sistemas computacionais heterogêneos paralelos/distribuídos. Uma ampla revisão bibliográfica com a correspondente análise crítica é apresentada. Essa revisão é a base para a comparação das métricas existentes para a avaliação do grau de heterogeneidade/homogeneidade dos sistemas computacionais. Uma nova métrica é proposta neste trabalho, removendo as restrições identificadas no estudo comparativo realizado. Resultados de aplicações dessa nova métrica são apresentados e discutidos. Esta tese propõe também o conceito de heterogeneidade/homogeneidade temporal que pode ser utilizado para futuros aprimoramentos de políticas de escalonamento empregadas em plataformas computacionais heterogêneas paralelas/distribuídas. Um novo índice de desempenho (Vector for Index of Performance - VIP), generalizando o conceito de índice de carga, é proposto com base em uma métrica Euclidiana. Esse novo índice é aplicado na implementação de uma política de escalonamento e amplamente testado através de modelagem e simulação. Os resultados obtidos são apresentados e analisados estatisticamente. É demonstrado que o novo índice leva a bons resultados de modo geral e é apresentado um mapeamento mostrando as vantagens e desvantagens de sua adoção quando comparado às métricas tradicionais. / This thesis approaches the problem of evaluating an adequate load index or a performance index, for using in process scheduling in heterogeneous parallel/distributed computing systems. A wide literature review with the corresponding critical analysis is presented. This review is the base for the comparison of the existing metrics for the evaluation of the computing systems homogeneity/heterogeneity degree. A new metric is proposed in this work, removing the restrictions identified during the comparative study realized. Results from the application of the new metric are presented and discussed. This thesis also proposes the concept of temporal heterogeneity/homogeneity that can be used for future improvements in scheduling polices for parallel/distributed heterogeneous computing platforms. A new performance index (Vector for Index of Performance - VIP), generalizing the concept of load index, is proposed based on an Euclidean metric. This new index is applied to the implementation of a scheduling police and widely tested through modeling and simulation. The results obtained are presented and statistically analyzed. It is shown that the new index reaches good results in general and it is also presented a mapping showing the advantages and disadvantages of its adoption when compared with the traditional metrics.