Em um ambiente de computação em nuvem, as empresas têm a capacidade de alocar recursos de acordo com a demanda. No entanto, há um atraso que pode levar alguns minutos entre o pedido de um novo recurso e o mesmo estar pronto para uso. Por esse motivo, as técnicas reativas, que solicitam um novo recurso apenas quando o sistema atinge um determinado limiar de carga, não são adequadas para o processo de alocação de recursos. Para resolver esse problema, é necessário prever as requisições que chegam ao sistema, no próximo período de tempo, para alocar os recursos necessários antes que o sistema fique sobrecarregado. Existem vários modelos de previsão de séries temporais para calcular as previsões de carga de trabalho com base no histórico de dados de monitoramento. No entanto, é difícil saber qual é o melhor modelo de previsão a ser utilizado em cada caso. A tarefa se torna ainda mais complicada quando o usuário não tem muitos dados históricos a serem analisados. A maioria dos trabalhos relacionados, considera apenas modelos de previsão isolados para avaliar os resultados. Outros trabalhos propõem uma abordagem que seleciona modelos de previsão adequados para um determinado contexto. Mas, neste caso, é necessário ter uma quantidade significativa de dados para treinar o classificador. Além disso, a melhor solução pode não ser um modelo específico, mas sim uma combinação de modelos. Neste trabalho propomos um método de previsão adaptativo, usando técnicas de otimização multiobjetivo, para combinar modelos de previsão de séries temporais. O nosso método não requer uma fase prévia de treinamento, uma vez que se adapta constantemente a medida em que os dados chegam ao sistema. Para avaliar a nossa proposta usamos quatro logs extraídos de servidores reais. Os resultados mostram que a nossa proposta frequentemente converge para o melhor resultado, e é suficientemente genérica para se adaptar a diferentes tipos de séries temporais. / In a cloud computing environment, companies have the ability to allocate resources according to demand. However, there is a delay that may take minutes between the request for a new resource and it is ready for using. The reactive techniques, which request a new resource only when the system reaches a certain load threshold, are not suitable for the resource allocation process. To address this problem, it is necessary to predict requests that arrive at the system in the next period of time to allocate the necessary resources, before the system becomes overloaded. There are several time-series forecasting models to calculate the workload predictions based on history of monitoring data. However, it is difficult to know which is the best time series forecasting model to be used in each case. The work becomes even more complicated when the user does not have much historical data to be analyzed. Most related work considers only single methods to evaluate the results of the forecast. Other work propose an approach that selects suitable forecasting methods for a given context. But in this case, it is necessary to have a significant amount of data to train the classifier. Moreover, the best solution may not be a specific model, but rather a combination of models. In this work we propose an adaptive prediction method using multi-objective optimization techniques to combine time-series forecasting models. Our method does not require a previous phase of training, because it constantly adapts the extent to which the data is coming. To evaluate our proposal we use four logs extracted from real servers. The results show that our proposal often brings the best result, and is generic enough to adapt to various types of time series.

Neste trabalho investiga-se a aplicação de um algoritmo genético multiobjetivo, ferramenta que se destaca por sua flexibilidade e interpretabilidade, na obtenção de fatores de ponderação para aplicação no controle preditivo de torque do motor de indução, ou Model Predictive Torque Control (MPTC). O MPTC busca minimizar a cada instante de atuação uma função custo que representa o sistema, destacando-se pela rápida resposta de torque, facilidade de incorporar restrições e ausência de modulador de tensão. No entanto, essa técnica apresenta fatores de ponderação em sua estrutura de cálculo que não dispõem de métodos analíticos de projeto. Utilizou-se o algoritmo genético de classificação nãodominada, ou Non-dominated Sorting Genectic Algorithm II (NSGA-II), projetado de forma a obter soluções que busquem o compromisso entre o desempenho dinâmico do motor, via minimização das oscilações de torque e fluxo, e a eficiência energética do sistema por meio da minimização da frequência média de chaveamento da eletrônica de potência. Resultados simulados e experimentais mostraram que o conjunto de soluções fornecido pelo NSGA-II é factível e contrapõe as oscilações de torque e de fluxo e a frequência média de chaveamento, cabendo à aplicação desejada a escolha da solução. Com isso, tem-se uma ferramenta de projeto dos fatores de peso do MPTC capaz de incorporar restrições e ajustar vários fatores ao mesmo tempo. / This work investigates the application of a multi-objective genetic algorithm to obtain a set of weighting factors suitable for use in Model Predictive Torque Control (MPTC) of a induction motor variable speed drive. MPTC approach aims at minimizing a cost function at each step, and is highlighted for its fast torque response, facility to incorporate system constraints and the absence of voltage modulators. Nevertheless, MPTC structure presents weighting factors in the cost function which lack of an analytical design procedure. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) was designed for a trade-off between torque and flux ripples minimization and minimization of the average switching frequency of the system. Simulated and experimental results showed NSGA-II offered a Pareto set of feasible solutions, so that torque ripple, flux ripple or average switching frequency can be minimized, depending on the solution chosen according to project demand. Thereby, there is a project tool for MPTC weighting factors able to adjust several factor at the same time, incorporating desired restrictions.

Multi-objective optimisation focuses on optimising multiple objectives simultanuously. Evolutionary and immune-based algorithms have been developed in order to solve multi-objective optimisation problems. These algorithms often include a property called elitism, a method of preserving good solutions. This study has focused on how different approaches of elitism affect an algorithm's ability to find optimal solutions in a multi-objective optimisation problem with a discrete and highly discontinuous decision space. Three state-of-the-art algorithms, NSGA-II, SPEA2+ and NNIA2, were implemented, validated and tested against a multi-objective optimisation problem of a miniature plant. Final populations yielded from all the algorithms were included in an analysis. The results of this study indicate that external populations are important in order for algorithms to find optimal solutions in multi-objective optimisation problems with a discrete and highly discontinuous decision spaces.

L’intégration de la recherche d’efficacité énergétique aux projets d’aménagement urbain est essentielle au vu du contexte actuel de transition énergétique et environnementale. Dans le but de réduire l’empreinte énergétique d’un quartier dès la phase de conception, un ensemble de contributions a été réalisé afin d’accompagner les aménageurs dans cette démarche. La plateforme de simulation DIMOSIM (DIstrict MOdeller and SIMulator) a été développée pour modéliser et simuler dynamiquement les flux énergétiques d’un quartier implanté au sein de son environnement urbain. La conception est optimisée à l’aide d’une procédure multiobjectif combinant les aspects énergétiques, économiques et environnementaux pour garantir la meilleure performance globale. Cette approche transversale est multi-étagée et intègre l’algorithme génétique NSGA-II afin de s’adapter aux spécificités du problème. La sélection de la solution préférentielle est ensuite facilitée par l’utilisation d’une méthode d’analyse multicritère de surclassement qui a été conçue dans le but de fournir une évaluation détaillée des différents concepts : la méthode ATLAS (Assistance TooL for decision support to Assess and Sort). Enfin, la procédure complète d’accompagnement a été appliquée à des projets de conception d’écoquartier pour en valider le fonctionnement mais également pour fournir l’aide à la décision nécessaire aux décideurs. / The integration of the research of energy efficiency into urban development projects is essential in the current context of energy and environmental transition. In order to reduce the energy footprint of a district already starting from the design phase, a set of contributions was elaborated to support the planners in this process. The DIMOSIM simulation platform (DIstrict MOdeller and SIMulator) has been developed to dynamically model and simulate the energy flows of a district located within its urban environment. The design of the district is optimized using a multi-objective procedure combining energy, economic and environmental aspects to ensure the best overall performance. A cross-cutting, multi-level approach integrating the NSGA-II genetic algorithm was implemented in order to adapt the procedure to the specificities of the problem. The selection of the preferred solution is then facilitated by the use of a multicriteria analysis method which was developed to provide a detailed evaluation of the different concepts : the outranking method ATLAS (Assistance TooL for decision support to assessment And Sort). Finally, the complete procedure dedicated to the district energy system design was applied to eco-district projects in order to validate its correct operation and also to provide the necessary support to decision-makers.

Energy and cost performance optimization for commercial building system design is growing in popularity, but it is often criticized for its time consuming process. Moreover, the current process lacks integration, which not only affects time performance, but also investors’ confidence in the predicted performance of the generated design. Such barriers keep building owners and design teams from embracing life cycle cost consideration. This thesis proposes a computationally efficient design optimization platform to improve the time performance and to streamline the workflow in an integrated multi-objective building system design optimization process. First, building system cost estimation is typically completed through a building information model based quantity take-off process, which does not provide sufficient design decision support features in the design process. To remedy this issue, an automatic cost estimation framework that integrates EnergyPlus with an external database to perform building systems’ capital and operation costs is proposed. Optimization, typically used for building system design selection, requires a large amount of computational time. The optimization process evaluates building envelope, electrical and HVAC systems in an integrated system not only to explore the cost-saving potential from a single high performance system, but also the interrelated effects among different systems. An innovative optimization strategy that integrates machine learning techniques with a conventional evolutionary algorithm is proposed. This strategy can reduce run time and improve the quality of the solutions. Lastly, developing baseline energy models typically takes days or weeks depending on the scale of the design. An automated system for generating baseline energy model according to ANSI/ASHRAE/IESNA Standard 90.1 performance rating method is thus proposed to provide a quick appraisal of optimal designs in comparison with the baseline energy requirements. The main contribution of this thesis is the development of a new design optimization platform to expedite the conventional decision making process. The platform integrates three systems: (1) cost estimation, (2) optimization and (3) benchmark comparison for minimizing the first cost and energy operation costs. This allows designers to confidently select an optimal design with high performance building systems by making a comparison with the minimum energy baseline set by standards in the building industry. Two commercial buildings are selected as case studies to demonstrate the effectiveness of this platform. One building is the Center for Sustainable Landscapes in Pittsburgh, PA. This case study is used as a new construction project. With 54 million possible design solutions, the platform is able to identify optimal designs in four hours. Some of the design solutions not only save the operation costs by up to 23% compared to the ASHRAE baseline design, but also reduce the capital cost ranging from 5% to 23%. Also, compared with the ASHRAE baseline design, one design solution demonstrates that the high investment of a product, building integrative photovoltaic (BiPV) system, can be justified through the integrative design optimization approach by the lower operation costs (20%) as well as the lower capital cost (12%). The second building is the One Montgomery Plaza, a large office building in Norristown, PA. This case study focuses on using the platform for a retrofit project. The calibrated energy model requires one hour to complete the simulation. There are 4000 possible design solutions proposed and the platform is able to find the optimal design solution in around 50 hours. Similarly, the results indicate that up to 25% capital cost can be saved with $1.7 million less operation costs in 25 years, compare to the ASHRAE baseline design.

L'adoption d'une méthodologie est cruciale pour la conception de procédés chimiques optimaux. L'optimisation multi-objective de modèles rigoureux en est un exemple, jouissant d'une application extensive dans la littérature. Cette méthode retourne un ensemble de solutions, dit de Pareto, présentant un compromis optimal entre les fonctions objectives. Ceci est suivi par une étape de sélection d'une solution d'intérêt répondant à des critères définis. Cette méthodologie s'appliquait, dans le cadre de cette thèse, à deux procédés. Le premier consistait en une distillerie d'éthanol à partir de la canne à sucre, combinée à un système de cogénération et de combustion à partir de la biomasse de canne à sucre. Le deuxième contenait en plus un système d'hydrolyse enzymatique de cette biomasse. Notre première contribution traitait de la construction d'une procédure pour la modélisation, simulation, intégration thermique et évaluation du coût des équipements. La deuxième contribution traitait de l'analyse des résultats réalisée à travers un suivi de variables mesurées, une fragmentation de la courbe de Pareto, une hiérarchisation des variables de décision et une comparaison avec la littérature. La dernière contribution traitait de l'étape de sélection qui s'est réalisée à travers une évaluation économique des solutions, sous des scénarii différents, avec la Valeur Nette Présente comme critère de sélection. En conclusion, cette thèse constitue une première application intégrale de la méthodologie proposée. Elle représente, de par ses contributions, un tremplin pour des applications futures à des procédés chimiques ou biochimiques, plus spécialement pour la canne à sucre / The use of a systematic methodology is crucial for the design of optimal chemical processes, namely bio-processes. Multi-objective optimization of rigorous process models is therein a prime example, with extensive use in literature. This method yields a Pareto set of optimal compromise solutions, from which one optimal solution is chosen based on specific criteria. This methodology was applied, in the course of this thesis, to two studied processes. The first consisted in a distillery converting sugarcane to ethanol, combined with a sugarcane biomass combustion and power cogeneration system. The second contained an additional biomass hydrolysis system. Our first contribution deals with the construction of an organized procedure for the modeling, simulation, heat integration and equipment and capital cost estimation of chemical processes. The second contribution deals with the analysis of the optimization results through a tracking of measured variables, the fragmentation of the Pareto curve, an ordering of optimization variables, and a comparisons with literature results. The final realization deals with the selection step realized through an economic evaluation of optimal solutions for various scenarios, with the Net Present Value as the selection criterion. In conclusion, this thesis constitutes a first integral application of the said methodology. It sets, through its contributions, a stepping stone for future application in the field of chemical and biochemical processes, namely for sugarcane processes

Environmentally conscious construction has received a significant amount of research attention during the last decades. Even though construction literature is rich in studies that emphasize the importance of environmental impact during the construction phase, most of the previous studies failed to combine environmental analysis with other project performance criteria in construction. This is mainly because most of the studies have overlooked the multi-objective nature of construction projects. In order to achieve environmentally conscious construction, multi-objectives and their relationships need to be successfully analyzed in the complex construction environment. The complex construction system is composed of changing project conditions that have an impact on the relationship between time, cost and environmental impact (TCEI) of construction operations. Yet, this impact is still unknown by construction professionals. Studying this impact is vital to fulfill multiple project objectives and achieve environmentally conscious construction. This research proposes an analytical framework to analyze the impact of changing project conditions on the relationship of TCEI. This study includes green house gas (GHG) emissions as an environmental impact category. The methodology utilizes multi-agent systems, multi-objective optimization, analytical network process, and system dynamics tools to study the relationships of TCEI and support decision-making under the influence of project conditions. Life cycle assessment (LCA) is applied to the evaluation of environmental impact in terms of GHG. The mixed method approach allowed for the collection and analysis of qualitative and quantitative data. Structured interviews of professionals in the highway construction field were conducted to gain their perspectives in decision-making under the influence of certain project conditions, while the quantitative data were collected from the Florida Department of Transportation (FDOT) for highway resurfacing projects. The data collected were used to test the framework. The framework yielded statistically significant results in simulating project conditions and optimizing TCEI. The results showed that the change in project conditions had a significant impact on the TCEI optimal solutions. The correlation between TCEI suggested that they affected each other positively, but in different strengths. The findings of the study will assist contractors to visualize the impact of their decision on the relationship of TCEI.

This doctoral thesis investigates the application of simulation-based optimization (SBO) as an alternative to conventional optimization techniques when the inherent uncertainty and complex features of real manufacturing systems need to be considered. Inspired by a real-world production planning setting, we provide a general formulation of the situation as an extended knapsack problem. We proceed by proposing a solution approach based on single and multi-objective SBO models, which use simulation to capture the uncertainty and complexity of the manufacturing system and employ meta-heuristic optimizers to search for near-optimal solutions. Moreover, we consider the design of matheuristic approaches that combine the advantages of population-based meta-heuristics with mathematical programming techniques. More specifically, we consider the integration of mathematical programming techniques during the initialization stage of the single and multi-objective approaches as well as during the actual search process. Using data collected from a manufacturing company, we provide evidence for the advantages of our approaches over conventional methods (integer linear programming and chance-constrained programming) and highlight the synergies resulting from the combination of simulation, meta-heuristics and mathematical programming methods. In the context of the same real-world problem, we also analyse different single and multi-objective SBO models for robust optimization. We demonstrate that the choice of robustness measure and the sample size used during fitness evaluation are crucial considerations in designing an effective multi-objective model.

A predição de estruturas de proteínas (Protein Structure Prediction PSP) é um problema computacionalmente complexo. Para tratar esse problema, modelos simplificados de proteínas, como o Modelo HP, têm sido empregados para representar as conformações e Algoritmos Evolutivos (AEs) são utilizados na busca por soluções adequadas para PSP. Entretanto, abordagens utilizando AEs muitas vezes não tratam adequadamente as soluções geradas, prejudicando o desempenho da busca. Neste trabalho, é apresentada uma formulação multiobjetivo para PSP em Modelo HP, de modo a avaliar de forma mais robusta as conformações produzidas combinando uma avaliação baseada no número de contatos hidrofóbicos com a distância entre os monômeros. Foi adotado o Algoritmo Evolutivo Multiobjetivo em Tabelas (AEMT) a fim de otimizar essas métricas. O algoritmo pode adequadamente explorar o espaço de busca com pequeno número de indivíduos. Como consequência, o total de avaliações da função objetivo é significativamente reduzido, gerando um método para PSP utilizando Modelo HP mais rápido e robusto / Protein Structure Prediction (PSP) is a computationally complex problem. To overcome this drawback, simplified models of protein structures, such as the HP Model, together with Evolutionary Algorithms (EAs) have been investigated in order to find appropriate solutions for PSP. EAs with the HP Model have shown interesting results, however, they do not adequately evaluate potential solutions by using only the usual metric of hydrophobic contacts, hamming the performance of the algorithm. In this work, we present a multi-objective approach for PSP using HP Model that performs a better evaluation of the solutions by combining the evaluation based on the number of hydrophobic contacts with the distance among the hydrophobic amino acids. We employ a Multi-objective Evolutionary Algorithm based on Sub-population Tables (MEAT) to deal with these two metrics. MEAT can adequately explore the search space with relatively low number of individuals. As a consequence, the total assessments of the objective function is significantly reduced generating a method for PSP using HP Model that is faster and more robust

Neste trabalho investiga-se a aplicação de um algoritmo genético multiobjetivo, ferramenta que se destaca por sua flexibilidade e interpretabilidade, na obtenção de fatores de ponderação para aplicação no controle preditivo de torque do motor de indução, ou Model Predictive Torque Control (MPTC). O MPTC busca minimizar a cada instante de atuação uma função custo que representa o sistema, destacando-se pela rápida resposta de torque, facilidade de incorporar restrições e ausência de modulador de tensão. No entanto, essa técnica apresenta fatores de ponderação em sua estrutura de cálculo que não dispõem de métodos analíticos de projeto. Utilizou-se o algoritmo genético de classificação nãodominada, ou Non-dominated Sorting Genectic Algorithm II (NSGA-II), projetado de forma a obter soluções que busquem o compromisso entre o desempenho dinâmico do motor, via minimização das oscilações de torque e fluxo, e a eficiência energética do sistema por meio da minimização da frequência média de chaveamento da eletrônica de potência. Resultados simulados e experimentais mostraram que o conjunto de soluções fornecido pelo NSGA-II é factível e contrapõe as oscilações de torque e de fluxo e a frequência média de chaveamento, cabendo à aplicação desejada a escolha da solução. Com isso, tem-se uma ferramenta de projeto dos fatores de peso do MPTC capaz de incorporar restrições e ajustar vários fatores ao mesmo tempo. / This work investigates the application of a multi-objective genetic algorithm to obtain a set of weighting factors suitable for use in Model Predictive Torque Control (MPTC) of a induction motor variable speed drive. MPTC approach aims at minimizing a cost function at each step, and is highlighted for its fast torque response, facility to incorporate system constraints and the absence of voltage modulators. Nevertheless, MPTC structure presents weighting factors in the cost function which lack of an analytical design procedure. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) was designed for a trade-off between torque and flux ripples minimization and minimization of the average switching frequency of the system. Simulated and experimental results showed NSGA-II offered a Pareto set of feasible solutions, so that torque ripple, flux ripple or average switching frequency can be minimized, depending on the solution chosen according to project demand. Thereby, there is a project tool for MPTC weighting factors able to adjust several factor at the same time, incorporating desired restrictions.