ESTiG - Artigos em Revistas Indexados à WoS/Scopus
URI permanente para esta coleção:
Navegar
Percorrer ESTiG - Artigos em Revistas Indexados à WoS/Scopus por Domínios Científicos e Tecnológicos (FOS) "Ciências Agrárias::Agricultura, Silvicultura e Pescas"
A mostrar 1 - 3 de 3
Resultados por página
Opções de ordenação
- Empowering olive cultivation with artificial intelligence: a systematic literature review on advancements and prospectsPublication . Mendes, João; Lima, José; Costa, Lino; Pereira, Ana I.This study provides a Systematic Literature Review on the application of Artificial Intelligence algorithms in the primary sector of olive cultivation. It compiles and analyses a collection of studies that leverage AI to enhance the efficiency and sustainability of olive production, maintenance, and harvesting processes. In this study, 43 papers were reviewed from the databases IEEE, Scopus, and Web of Science through the Preferred Reporting Items for Systematic Reviews and Meta-Analyses method. This research aims to identify AI applications in the primary olive growing sector. The findings highlight a significant trend toward adopting advanced AI techniques, particularly Deep Learning algorithms such as Convolutional Neural Networks, for many tasks ranging from cultivar identification and foliar disease classification to crop yield forecasting with high accuracies.
- Impact of hyper-parameter tuning on CNN accuracy in agricultural image classificationPublication . Mendes, João; Lima, José; Costa, Lino; Hendrix, Eligius M.T.; Pereira, Ana I.This study explores the impact of hyper-parameter optimization on the performance of convolutional neural networks (CNNs) for olive cultivar classification using transfer learning. Pre-trained ImageNet models such as VGG16, InceptionV3, and ResNet50 were adapted to a proprietary dataset, with VGG16 selected for detailed evaluation. Key hyper-parameters, including layer count, neurons per layer, dropout rate, learning rate, and batch size, were tuned using random search. The best configuration achieved a validation accuracy of 87.5%, significantly outperforming the control model. Sensitivity analyses with Morris and Sobol methods identified the number of layers as the most influential factor, followed by dropout and learning rates through interaction effects. These findings demonstrate the importance of tailoring CNN architecture and regularization settings to the problem domain. These results underscore the importance of tuning architectural depth and regularization mechanisms for performance optimization. As a practical guideline, models with fewer layers and intermediate dropout levels demonstrated higher robustness and generalization, offering an effective strategy for adapting CNNs to agricultural classification tasks.
- On-site power generation using biogas in sewage treatment plants: a techno-economic assessment of a brazilian uasb facilityPublication . Pérez, Nestor Proenza; Adrião Cabral, Edilson; Bimestre, Thiago Averaldo; Loures, Carla Almeida; Yepes Maya, Diego M.; Ribeiro, Luís FrölénSmall sewage treatment plants (STPs) in developing regions often flare the biogas produced in their upflow anaerobic sludge blanket (UASB) reactors, giving away a cost-effective energy source. This study tests whether on-site biogas-to-energy can pay for itself in approximately 2 years, even in plants treating less than 30 l s(-1). A small-scale STP in Angra dos Reis, Brazil (25 L/s), was studied, with an average biogas flow of 9.7 m(3)/h; electricity generation was modeled for an engine generator unit with an efficiency of 30%. The techno-economic results show that the actual system would generate 125 MWh/year at a levelized cost of 0.017-0.023 USD/kWh, covering 47% of the plant's electricity demand. At a discount rate of 8%, the net present value was + 9.3 k US$, and the simple payback period was 2 years for the initial investment. Additionally, extrapolating the results to account for future expansion of the sewage treatment plant based on the total population in the region served by the system reveals even more promising results, with a suggested payback period of 1 year and 1 month of operation, covering approximately 57% of electricity demand. Scaling this retrofit to the approximately 18,000 comparable UASB-based STPs worldwide at low capital cost could reduce electricity bills by approximately 40% and avoid similar to 450 tons of CO2-eq. per plant per year through methane capture and displacement of electricity from the grid. These results confirm that decentralized biogas power generation on a small scale is not only technically feasible, but also financially attractive and ecologically beneficial for operators of sewage and wastewater treatment plants and municipalities.