The effect of demand information sharing in a supply chain under demand uncertainty: a simulation study

Dissertação para obtenção do Grau de Mestre em Engenharia e Gestão Industrial

Gas adsorption in the MIL-53(AI) metal organic framework. Experiments and molecular simulation

Dissertação para obtenção do Grau de Doutor em Engenharia Química

A simulation model for lean, agile, resilient and green supply chain management: practices and interoperability assessment

Dissertação para obtenção do Grau de Mestre em Engenharia e Gestão Industrial

Analysis and simulation of social unrest in Europe: towards understanding social unrest in Europe

Dissertation submitted in partial fulfillment of the requirements for the Degree of Master of Science in Geospatial Technologies

Liquidity, business cycles and monetary policy: A simulation

A Masters Thesis, presented as part of the requirements for the award of a Research Masters Degree in Economics from NOVA – School of Business and Economics

A system’s approach to cache hierarchy-aware decomposition of data-parallel computations

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Optimization of breast tomosynthesis image reconstruction using parallel computing

Breast cancer is the most common cancer among women, being a major public health problem. Worldwide, X-ray mammography is the current gold-standard for medical imaging of breast cancer. However, it has associated some well-known limitations. The false-negative rates, up to 66% in symptomatic women, and the false-positive rates, up to 60%, are a continued source of concern and debate. These drawbacks prompt the development of other imaging techniques for breast cancer detection, in which Digital Breast Tomosynthesis (DBT) is included. DBT is a 3D radiographic technique that reduces the obscuring effect of tissue overlap and appears to address both issues of false-negative and false-positive rates. The 3D images in DBT are only achieved through image reconstruction methods. These methods play an important role in a clinical setting since there is a need to implement a reconstruction process that is both accurate and fast. This dissertation deals with the optimization of iterative algorithms, with parallel computing through an implementation on Graphics Processing Units (GPUs) to make the 3D reconstruction faster using Compute Unified Device Architecture (CUDA). Iterative algorithms have shown to produce the highest quality DBT images, but since they are computationally intensive, their clinical use is currently rejected. These algorithms have the potential to reduce patient dose in DBT scans. A method of integrating CUDA in Interactive Data Language (IDL) is proposed in order to accelerate the DBT image reconstructions. This method has never been attempted before for DBT. In this work the system matrix calculation, the most computationally expensive part of iterative algorithms, is accelerated. A speedup of 1.6 is achieved proving the fact that GPUs can accelerate the IDL implementation.

Molecular simulation of gas adsorption equilibria in nanoporous materials

This work is divided into two distinct parts. The first part consists of the study of the metal organic framework UiO-66Zr, where the aim was to determine the force field that best describes the adsorption equilibrium properties of two different gases, methane and carbon dioxide. The other part of the work focuses on the study of the single wall carbon nanotube topology for ethane adsorption; the aim was to simplify as much as possible the solid-fluid force field model to increase the computational efficiency of the Monte Carlo simulations. The choice of both adsorbents relies on their potential use in adsorption processes, such as the capture and storage of carbon dioxide, natural gas storage, separation of components of biogas, and olefin/paraffin separations. The adsorption studies on the two porous materials were performed by molecular simulation using the grand canonical Monte Carlo (μ,V,T) method, over the temperature range of 298-343 K and pressure range 0.06-70 bar. The calibration curves of pressure and density as a function of chemical potential and temperature for the three adsorbates under study, were obtained Monte Carlo simulation in the canonical ensemble (N,V,T); polynomial fit and interpolation of the obtained data allowed to determine the pressure and gas density at any chemical potential. The adsorption equilibria of methane and carbon dioxide in UiO-66Zr were simulated and compared with the experimental data obtained by Jasmina H. Cavka et al. The results show that the best force field for both gases is a chargeless united-atom force field based on the TraPPE model. Using this validated force field it was possible to estimate the isosteric heats of adsorption and the Henry constants. In the Grand-Canonical Monte Carlo simulations of carbon nanotubes, we conclude that the fastest type of run is obtained with a force field that approximates the nanotube as a smooth cylinder; this approximation gives execution times that are 1.6 times faster than the typical atomistic runs.

Agent-based parking occupancy simulation

The existing parking simulations, as most simulations, are intended to gain insights of a system or to make predictions. The knowledge they have provided has built up over the years, and several research works have devised detailed parking system models. This thesis work describes the use of an agent-based parking simulation in the context of a bigger parking system development. It focuses more on flexibility than on fidelity, showing the case where it is relevant for a parking simulation to consume dynamically changing GIS data from external, online sources and how to address this case. The simulation generates the parking occupancy information that sensing technologies should eventually produce and supplies it to the bigger parking system. It is built as a Java application based on the MASON toolkit and consumes GIS data from an ArcGis Server. The application context of the implemented parking simulation is a university campus with free, on-street parking places.

Numerical simulation of blast effects on fibre grout RC panels

The main purpose of the present dissertation is the simulation of the response of fibre grout strengthened RC panels when subjected to blast effects using the Applied Element Method, in order to validate and verify its applicability. Therefore, four experimental models, three of which were strengthened with a cement-based grout, each reinforced by one type of steel reinforcement, were tested against blast effects. After the calibration of the experimental set-up, it was possible to obtain and compare the response to the blast effects of the model without strengthening (reference model), and a fibre grout strengthened RC panel (strengthened model). Afterwards, a numerical model of the reference model was created in the commercial software Extreme Loading for Structures, which is based on the Applied Element Method, and calibrated to the obtained experimental results, namely to the residual displacement obtained by the experimental monitoring system. With the calibration verified, it is possible to assume that the numerical model correctly predicts the response of fibre grout RC panels when subjected to blast effects. In order to verify this assumption, the strengthened model was modelled and subjected to the blast effects of the corresponding experimental set-up. The comparison between the residual and maximum displacements and the bottom surface’s cracking obtained in the experimental and the numerical tests yields a difference of 4 % for the maximum displacements of the reference model, and a difference of 4 and 10 % for the residual and maximum displacements of the strengthened model, respectively. Additionally, the cracking on the bottom surface of the models was similar in both methods. Therefore, one can conclude that the Applied ElementMethod can correctly predict and simulate the response of fibre grout strengthened RC panels when subjected to blast effects.

A public transport bus assignment problem: parallel metaheuristics assessment

Combinatorial Optimization Problems occur in a wide variety of contexts and generally are NP-hard problems. At a corporate level solving this problems is of great importance since they contribute to the optimization of operational costs. In this thesis we propose to solve the Public Transport Bus Assignment problem considering an heterogeneous fleet and line exchanges, a variant of the Multi-Depot Vehicle Scheduling Problem in which additional constraints are enforced to model a real life scenario. The number of constraints involved and the large number of variables makes impracticable solving to optimality using complete search techniques. Therefore, we explore metaheuristics, that sacrifice optimality to produce solutions in feasible time. More concretely, we focus on the development of algorithms based on a sophisticated metaheuristic, Ant-Colony Optimization (ACO), which is based on a stochastic learning mechanism. For complex problems with a considerable number of constraints, sophisticated metaheuristics may fail to produce quality solutions in a reasonable amount of time. Thus, we developed parallel shared-memory (SM) synchronous ACO algorithms, however, synchronism originates the straggler problem. Therefore, we proposed three SM asynchronous algorithms that break the original algorithm semantics and differ on the degree of concurrency allowed while manipulating the learned information. Our results show that our sequential ACO algorithms produced better solutions than a Restarts metaheuristic, the ACO algorithms were able to learn and better solutions were achieved by increasing the amount of cooperation (number of search agents). Regarding parallel algorithms, our asynchronous ACO algorithms outperformed synchronous ones in terms of speedup and solution quality, achieving speedups of 17.6x. The cooperation scheme imposed by asynchronism also achieved a better learning rate than the original one.

MarketOpolis: A market simulation based on investors decisions

This study aims to replicate Apple’s stock market movement by modeling major investment profiles and investors. The present model recreates a live exchange to forecast any predictability in stock price variation, knowing how investors act when it concerns investment decisions. This methodology is particularly relevant if, just by observing historical prices and knowing the tendencies in other players’ behavior, risk-adjusted profits can be made. Empirical research made in the academia shows that abnormal returns are hardly consistent without a clear idea of who is in the market in a given moment and the correspondent market shares. Therefore, even when knowing investors’ individual investment profiles, it is not clear how they affect aggregate markets.