Teses de Mestrado ESTiG
Permanent URI for this community
Browse
Browsing Teses de Mestrado ESTiG by Sustainable Development Goals (SDG) "13:Ação Climática"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
- A deep learning system for daily activity recognition in smart home environmentsPublication . Dessanti, Augusto Luvisa; Lima , JoséThis work presents the implementation of a system for daily activity classification using 3D Convolutional Neural Networks (3D CNN) and the Toyota Smarthome Dataset. The system aims to generalize and correctly classify activities, even in the face of data limitations such as high class imbalance, ambient occlusions, and similarities between classes. To overcome these challenges, preprocessing techniques and data augmentation were applied, including spatio-temporal resizing and image enhancement, with the objective of optimizing learning and generalization capabilities of the model. The proposed approach proved to be effective compared to other models on the same dataset, achieving 85.7% accuracy, 0.8568 precision, and 0.8570 recall.
- Assessment of carbon sequestration in forest areas using deep learningPublication . Britto, Raphael Duarte; Lima, José; Pereira, Ana I.; Santos, Murillo Ferreira dosGrowing awareness of environmental impacts is making it more important than ever to explore regions with dense vegetation. Remote monitoring is a viable solution for the surveillance of large areas, such as forests. Based in intelligent systems, this work aims to develop a methodology for assessing carbon sequestration in forest areas. Deep learning (DL) structures were used to predict the heights and stand densities in tree colonies. Light Detection and Ranging (LiDAR) sensor scans obtained by Unmanned Aerial Vehicle (UAV) overflight were processed to extract elevation values and images. Point clouds were processed using QGIS software. The LAStools extension was employed to manipulate Digital Elevation Model (DEM) and rasters, obtaining relevant information. This data was then used to create a dataset for implementation in Convolutional Neural Network (CNN) models. Specific biometric relationships were implemented to estimate additional data such as Above Ground Biomass (AGB) and phytovolume. After evaluating different architectures, the VGG19 CNN model was highlighted as the most promising. An area of 46.6 hectares was covered, with an estimated total value of 4225.81 tons of carbon. This value provided an accuracy of 91%, based on forest inventories carried out in the same region. The study was conducted in the northern region of mainland Portugal, encompassing two distinct Pinus pinaster Ait. forests.
- Decarbonizing Brazil’s power sector: high-resolution simulation and lifecycle emissions analysis of a 100% renewable gridPublication . Roma, Gabriel Fonseca Oliveira; Ferreira, Ângela P.; Dranka, Géremi GilsonÀ medida que o mundo acelera os esforços para combater as alterações climáticas, a transição para sistemas eléctricos totalmente renováveis tornou-se um objectivo crucial, especialmente para economias emergentes. Este trabalho explora a viabilidade de um sistema eléctrico 100% renovável no Brasil até 2050, através de simulações de alta resolução utilizando o modelo EnergyPLAN e uma Avaliação do Ciclo de Vida (ACV) probabilística das tecnologias renováveis consideradas. Ferramentas tradicionais de planeamento energético e estratégias nacionais de longo prazo tendem a subestimar ou ignorar as emissões indirectas de gases com efeito de estufa (GEE) associadas ao ciclo de vida das tecnologias renováveis — incluindo construção, fabrico, manutenção e desmantelamento —, que podem alterar significativamente o perfil ambiental dos cenários futuros. Assim, são necessárias abordagens integradas de modelação que combinem simulações detalhadas com análises ambientais robustas. A metodologia adoptada inclui modelação operacional horária baseada nas projecções do Plano Nacional de Energia (PNE 2050) e a construção de um Ano Meteorológico Típico (TMY) para representar perfis realistas de geração. Foi elaborado um inventário detalhado de emissões indirectas de GEE para centrais hidroeléctricas, eólicas, solares fotovoltaicas, biomassa e nucleares, complementado por uma simulação de Monte Carlo para capturar incertezas. Os resultados indicam que uma rede 100% renovável poderá ainda emitir, em média, 30,8 MtCO2eq/ano devido às emissões indirectas. A inclusão de sistemas combinados de armazenamento energético poderá reduzir as necessidades de importação de electricidade de 19,87 TWh para 8,8 TWh, aumentando a capacidade do sistema para lidar com a variabilidade sazonal. Estes resultados reforçam a viabilidade técnica da transição e evidenciam a importância do investimento estratégico em armazenamento e da contabilização ambiental rigorosa.
- Treatment of textile wastewater using electrochemical oxidationPublication . Jebali, Sarra; Peres, António M.; Fajardo, Ana; Veloso, Ana; Hamrouni, AbdessalemWater pollution is a significant environmental issue, driven by contaminants from industrial, agricultural, and domestic sources. Recently, persistent organic compounds have emerged as major pollutants, prompting the need for more advanced treatment solutions. Although traditional methods such as filtration, adsorption, and biological processes have been applied, they often suffer from low efficiency, high energy costs, and the risk of secondary pollution. Therefore, electrochemical methods have gained attention as efficient alternatives, offering controlled and verifiable oxidation and reduction reactions for pollutant degradation, with the added benefits of versatility and minimal secondary waste. In this study, electrochemical oxidation was implemented to treat water pollutants, evaluating the performance of different anode materials boron-doped diamond (BDD), titanium coated with iridium dioxide (Ti/IrO₂), and titanium coated with ruthenium dioxide (Ti/RuO₂). In addition to electrode material, the effects of several key operational parameters were investigated, including current density, inter-electrode distance, mixing rate, and initial dye concentration, in order to determine their impact on color and chemical oxygen demand (COD) removal efficiencies. The BDD anode demonstrated superior performance, achieving 100% color removal consistently across all current densities, presenting optimal results. A current density of 0.06 A/cm² was selected as it offers ideal balance in terms of color and COD removal efficiency as well as cost-effectiveness, with COD removal reaching 30.3%. The operational parameters were systematically optimized to enhance the efficiency of the electrochemical oxidation process, and an inter-electrode distance of 0.5 cm was found to be the most effective, yielding the highest COD removal of 35.7 ± 0.56 (95% CI: [35.6, 35.9]; a mixing rate of 250 rpm led to a COD removal of 68.8 ± 0.67 (95% CI: [68.7, 68.9]); and an initial methylene blue concentration of 50 ppm resulted in the highest COD removal, reaching 80.9 ± 0.03 (95% CI: [80.7, 81.1]). These findings emphasize the potential of electrochemical oxidation, particularly using BDD anodes under optimized conditions, for efficient and sustainable water treatment.