Universitat Politècnica de Catalunya
UPCommons. Portal del coneixement obert de la UPCNot a member yet
290727 research outputs found
Sort by
Evaluación de riesgo por tormentas eléctricas en energías renovables utilizando datos satelitales
No full textAquest Treball de Fi de Grau analitza el risc associat a les tempestes elèctriques en sistemes d’energia renovable, com instal·lacions solars i eòliques, mitjançant l’ús de dades satel·litals. S’estudia la distribució espacial i temporal de l’activitat elèctrica per identificar zones de risc més alt. Els resultats permeten donar suport a la planificació i protecció d’infraestructures renovables davant de descàrregues elèctriques.El objeto de este trabajo tiene como finalidad el desarrollo de metodologías para evaluar el riesgo por tormentas eléctricas en plantas fotovoltaicas en España. En los últimos años, se ha evidenciado un aumento en fenómenos meteorológicos adversos, estos hechos requieren la necesidad de investigar más a fondo cómo las tormentas eléctricas atmosféricas afectan las infraestructuras energéticas. Dado que la energía solar es una de las principales fuentes de electricidad, y la principal de las energías renovables en el país, es crucial asegurar la protección y el buen funcionamiento de estas instalaciones para mantener la continuidad y estabilidad del suministro energético nacional. Para el desarrollo de esta investigación se ha empleado las nuevas tecnologías de información de tormentas eléctricas y rayos derivadas de satelitales, las cuales se vienen posicionando como herramientas novedosas y altamente tecnológicas para identificar, monitorear y realizar análisis predictivos en las actividades de operación y mantenimiento de plantas fotovoltaicas. Este estudio propone una metodología para identificar factores de riesgo asociados con tormentas eléctricas en instalaciones solares en España basados en la información del satélite meteorológico Meteosat Third Generation Lightning Imager (MTG-LI) de EUTMETSAT. Con el conjunto de datos obtenidos de este satélite, se llevará a cabo un análisis espacial y temporal de las descargas eléctricas a través de bases de datos espaciales que permitirá gestionar de manera eficiente el gran volumen de información que proporcionan la fuente de datos satelital. Finalmente, se generarán mapas temáticos y representaciones cartográficas que ayudarán a visualizar y comprender los riesgos identificados, ofreciendo un soporte gráfico claro y accesible para la toma de decisiones en la protección y planificación de las plantas fotovoltaicas.This work focuses on the development of methodologies to assess the risk posed by thunderstorms to photovoltaic plants in Spain. In recent years, an increase in adverse meteorological phenomena has been observed, highlighting the need to further investigate how atmospheric thunderstorms affect energy infrastructures. Given that solar energy is one of the country’s main sources of electricity, it is crucial to ensure the protection and proper functioning of these facilities to maintain the continuity and stability of the national energy supply. To carry out this research, new technologies for thunderstorm and lightning information derived from satellite data have been employed. These technologies are becoming established as innovative and highly advanced tools for identifying, monitoring, and performing predictive analyses in the operation and maintenance activities of photovoltaic and solar plants. This study proposes a methodology to identify risk factors associated with thunderstorms in solar installations in Spain, based on data from the Meteosat Third Generation Lightning Imager (MTG-LI) meteorological satellite operated by EUMETSAT. Using this dataset, a spatial and temporal analysis of lightning discharges will be conducted through spatial databases, enabling efficient management of the large volume of information provided by the satellite data source. Finally, thematic maps and cartographic representations will be generated to help visualize and understand the identified risks, providing clear and accessible graphical support for decision-making in the protection and planning of photovoltaic plants
Migración de SQL Server 2016-2022
No full textAquest Treball de Fi de Grau presenta la planificació, coordinació i validació d'una migració d'infraestructura de bases de dades des de Microsoft SQL Server 2016 cap a SQL Server 2022 dins d'una organització del sector públic. L'objectiu principal del projecte és garantir la continuïtat del servei, millorar la seguretat i mantenir el compliment normatiu, minimitzant al mateix temps els riscos operatius associats a la migració de sistemes de bases de dades crítics per al negoci. El treball se centra en l'anàlisi de la infraestructura existent, la definició de l'estratègia de migració i el suport a les activitats de migració en entorns no productius (DEV i PRE). El desplegament final en producció queda explícitament fora de l'abast d'aquest projecte. Es posa un èmfasi especial en la gestió de riscos, l'anàlisi de dependències, els requisits d'infraestructura i seguretat, i la coordinació amb múltiples actors tècnics i organitzatius. Els resultats mostren que un enfocament estructurat i basat en el risc és essencial per a les migracions de bases de dades en entorns del sector públic regulats, on les restriccions administratives, els requisits de seguretat i les dependències externes tenen un paper clau en l'execució del projecte.This Bachelor's Thesis presents the planning, coordination, and validation of a database infrastructure migration from Microsoft SQL Server 2016 to SQL Server 2022 within a public-sector organization. The primary objective of the project is to ensure service continuity, improve security, and maintain regulatory compliance while minimizing the operational risks associated with the migration of business-critical database systems. The work focuses on the analysis of the existing infrastructure, the definition of a migration strategy, and the support of migration activities in non-production environments (DEV and PRE). The final production deployment is explicitly excluded from the scope of this project. Particular emphasis is placed on risk management, dependency analysis, infrastructure and security prerequisites, and coordination with multiple technical and organizational stakeholders. The results demonstrate that a structured, risk-driven approach is essential for database migrations in regulated public-sector environments, where administrative processes, security constraints, and external dependencies play a significant role in project execution.Este Trabajo de Fin de Grado presenta la planificación, coordinación y validación de una migración de infraestructura de bases de datos desde Microsoft SQL Server 2016 a SQL Server 2022 dentro de una organización del sector público. El objetivo principal del proyecto es garantizar la continuidad del servicio, mejorar la seguridad y mantener el cumplimiento normativo, minimizando al mismo tiempo los riesgos operativos asociados a la migración de sistemas de bases de datos críticos para el negocio. El trabajo se centra en el análisis de la infraestructura existente, la definición de la estrategia de migración y el soporte a las actividades de migración en entornos no productivos (DEV y PRE). El despliegue final en producción queda explícitamente fuera del alcance de este proyecto. Se pone un énfasis especial en la gestión de riesgos, el análisis de dependencias, los requisitos de infraestructura y seguridad, y la coordinación con múltiples actores técnicos y organizativos. Los resultados muestran que un enfoque estructurado y basado en el riesgo es esencial para las migraciones de bases de datos en entornos del sector público regulados, donde las restricciones administrativas, los requisitos de seguridad y las dependencias externas desempeñan un papel clave en la ejecución del proyecto
Exploring the use of prompts and AI-generated images to elicit explanations in engineering education
No full textLearning structural engineering concepts often presents challenges due to their abstract nature and the limitations of traditional teaching methods. One way to assess and develop conceptual understanding is through students' ability to generate high-quality explanations. While visualization tools are known to support this process, generative AI offers a novel approach for creating customized visual representations that may deepen student learning. This study investigates the potential of using generative AI tools—specifically, textual prompts and AI-generated images—to elicit and improve undergraduate students' explanations of key structural engineering concepts. Guided by two research questions, the study explores how these tools influence explanation quality and how the process of generating and reflecting on AI-produced images supports changes in conceptual understanding. Conducted in a steel structures design course with 29 students, the study asked participants to write initial explanations, generate images using AI tools, and then revise their explanations. Explanation quality was assessed using the SOLO taxonomy. Results showed that six of ten students who completed all stages advanced to higher SOLO levels, indicating improved conceptual depth. Students with lower initial understanding demonstrated the greatest improvement, while those with stronger prior knowledge experienced limited gains. The findings suggest that AI-generated images, when combined with structured guidance and clear instructional prompts, can support students in bridging abstract engineering concepts with concrete representations. However, visual tools alone are insufficient. This study emphasizes the need for intentional pedagogical design when integrating generative AI in engineering education and highlights future research opportunities to extend these approaches across diverse STEM contexts.The authors acknowledge the support of their affiliated institutions, whose resources were instrumental in the successful completion of this research. The authors also extend their gratitude to the students who participated in this research project for their invaluable contributions and active engagement throughout the study.Peer Reviewed4 - Educació de QualitatPostprint (author's final draft
Grup d'innovació DIACIM: disseny d'activitats científiques d'indagació i modelització
No full textEl grup d’innovació DIACIM, vinculat a l’ICE de la UAB i hereu del grup DIATIC, reuneix professorat de secundària i investigadors en didàctica de les ciències per fomentar pràctiques d’indagació i modelització científica entre l’alumnat d’ESO i Batxillerat. Durant el curs 2024-25, el grup ha treballat en tres nodes temàtics: “Arranquem”, centrat en com iniciar situacions d’aprenentatge des del constructivisme amb un decàleg de recomanacions; “Model Planeta Terra”, amb una proposta de conceptualització sistèmica dels grans sistemes terrestres i les seves interaccions; i “Llum a la foscor”, que presenta reptes d’investigació criminal per comprendre principis d’òptica a través d’experiments. Aquest treball col·laboratiu es desenvolupa en trobades mensuals i espais de formació i debat a la UAB
Wind Measurement Technologies: In-Situ and Remote Sensing
No full textTema 42025/20261r quadrimestrev.
Sistema agentic per avaluació de pràctiques i exercicis
No full textProblema Avui dia, l'avaluació d'activitats en plataformes d'aprenentatge com Moodle requereix una intervenció manual intensiva per part del professorat, i això limita la rapidesa de correcció i la personalització de la retroalimentació als estudiants. Tot i l'emergència de la intel·ligència artificial generativa com a eina àmpliament usada en la nostra societat, la seva integració amb sistemes de gestió de l'aprenentatge per a automatitzar processos educatius encara presenta reptes tècnics, tant en l'àmbit d'interoperabilitat com de seguretat de dades i sobretot de mantenibilitat i escalabilitat. Solució proposada Aquest projecte pretén desenvolupar un servidor que empri el Model Context Protocol (MCP) per a actuar com a pont entre agents d'intel·ligència artificial i l'API externa de la plataforma Moodle. El servidor implementa un conjunt extensiu de funcionalitats de l'API REST de webservices de Moodle com a eines MCP, que permet que agents d'IA puguin accedir a informació de cursos, estudiants, activitats i qualificació, i també publicar retroalimentació automatitzada i resultats d'avaluació. Metodologia S'implementa el servidor amb una arquitectura per capes que separa les responsabilitats del servidor MCP (exposició d'eines a l'agent IA), els models de validació de dades i el client de comunicació amb Moodle. S'ha escollit un enfocament de stateless proxy que garanteix seguretat, simplicitats i mantenibilitat. El projecte s'ha desenvolupat usant Python 3.11+, el framework FastMCP d'Anthropic, i Docker per a l'entorn de proves amb Moodle 5.1+ Resultats El servidor implementa 24 funcionalitats de Moodle organitzades en categories de gestió de cursos, usuaris, matriculacions i qualificacions. S'han validat mitjançant un conjunt de tests unitaris amb cobertura superior al 80% i proves d'integració tan semimanuals amb MCP Inspector com end-to-end amb l'agent IA Claude Desktop. El projecte aporta una solució reutilitzable i de codi obert que facilita l'ús d'IA en el camp de l'educació, reduint el temps i esforç necessari per a activitats educatives i fomentant l'ús d'eines d'intel·ligència artificial generativa per part d'institucions educatives.Problem Currently, assessing activities on learning platforms like Moodle requires intensive manual intervention from teachers, which limits the speed of grading and the personalization of feedback for students. Despite the emergence of generative artificial intelligence as a widely used tool in our society, its integration with learning management systems to automate educational processes still presents technical challenges, both in the areas of interoperability and data security, and above all, maintainability and scalability. Proposed solution This project aims to develop a server that uses the Model Context Protocol (MCP) to act as a bridge between artificial intelligence agents and the Moodle platform's external API. The server implements an extensive set of functionalities from Moodle's web services REST API as MCP tools, enabling AI agents to access information about courses, students, activities, and grades, as well as publish automated feedback and assessment results. Methodology The server is implemented with a layered architecture that separates the responsibilities of the MCP server (exposing tools to the AI ¿¿agent), the data validation models, and the client for communication with Moodle. A stateless proxy approach was chosen to ensure security, simplicity, and maintainability. The project was developed using Python 3.11+, the Anthropic FastMCP framework, and Docker for the test environment with Moodle 5.1+. Results The server implements 24 Moodle functionalities organized into categories for course management, users, enrollments, and grades. These have been validated through a set of unit tests with over 80% coverage and integration tests, ranging from semi-manual testing with MCP Inspector to end-to-end testing with the Claude Desktop AI agent. The project provides a reusable, open-source solution that facilitates the use of AI in education, reducing the time and effort required for educational activities and promoting the use of generative artificial intelligence tools by educational institutions.Problema Hoy en día, la integración de agentes de inteligencia artificial con plataformas de aprendizaje como Moodle representa retos técnicos de interoperabilidad, especialmente para tareas como la evaluación automatizada y la personalización de la retroalimentación a los estudiantes. A pesar de la emergencia de la inteligencia artificial generativa como herramienta ampliamente usada en nuestra sociedad, su integración con sistemas de gestión del aprendizaje para automatizar procesos educativos presenta todavía retos técnicos, tanto en el ámbito de interoperabilidad como de seguridad de datos y sobre todo de mantenibilidad y escalabilidad. Solución propuesta Este proyecto pretende desarrollar un servidor que utilice el Modelo Context Protocol (MCP) para actuar como puente entre agentes de inteligencia artificial y la API externa de la plataforma Moodle. El servidor implementa un conjunto extensivo de funcionalidades de la API REST de webservices de Moodle como herramientas MCP, que permite que agentes de IA puedan acceder a información de cursos, estudiantes, actividades y calificación, así como publicar retroalimentación automatizada y resultados de evaluación. Metodología Se implementa el servidor con una arquitectura por capas que separa las responsabilidades del servidor MCP (exposición de herramientas al agente IA), los modelos de validación de datos y el cliente de comunicación con Moodle. Se ha escogido un enfoque de stateless proxy que garantiza seguridad, simplicidades y mantenibilidad. El proyecto se ha desarrollado usando Python 3.11+, el framework FastMCP de Anthropic, y Docker para el entorno de pruebas con Moodle 5.1+. Resultados El servidor implementa 24 funcionalidades de Moodle organizadas en categorías de gestión de cursos, usuarios, matriculaciones y calificaciones. Se han validado mediante un conjunto de tests unitarios con cobertura superior al 80% y pruebas de integración tanto semimanuales como MCP Inspector como end-to-end con el agente IA Claude Desktop. El proyecto aporta una solución reutilizable y de código abierto que facilita el uso de IA en el campo de la educación, reduciendo el tiempo y el esfuerzo necesario para actividades educativas y fomentando el uso de herramientas de inteligencia artificial generativa por parte de instituciones educativas
Optimizing the operation of energy islands with predictive nonlinear programming: a case study based on the Princess Elisabeth Energy Island
No full textThe concepts of energy islands or energy hubs have gained attention in Europe as a means to enhance offshore wind integration and regional energy systems. These islands can incorporate high-voltage alternating current (HVAC) and high-voltage direct current (HVDC) transmission systems, battery energy storage systems (BESS), and hydrogen production, requiring advanced operational strategies to manage the inherent nonlinearities and time dependence of their subsystems. To address these challenges, this work proposes a comprehensive framework for the optimal operation of hybrid AC/DC energy islands, addressing (i) active and reactive power dispatch, incorporating BESS and hydrogen production; (ii) a detailed wind resource characterization based on 1 year of hourly data obtained using a realistic wind model with local measurements, including wake losses and turbine-level forecasts, used to define representative seasonal and spatial production patterns that inform typical operating conditions; (iii) operational optimization of a realistic test system based on the Princess Elisabeth Energy Island, and (iv) uncertainty analysis via Monte Carlo simulations, quantifying the impact of wind power and electricity price forecast errors, set up using commercial wind power planning tools and advanced forecasting software, and verified with Pyomo/Python.This work has received funding from the ADOreD project of the European Union’s Horizon Europe Research and Innovation program under the Marie Skłodowska-Curie grant agreement no. 101073554. This research has been supported by the Horizon Europe Marie Skłodowska-Curie Actions (grant no. 101073554). The work of Oriol Gomis-Bellmunt was supported by AGAUR – ICREA Academia program. The work of Marc CheahMane was supported by the Serra Hunter Program.Peer ReviewedPostprint (published version
Role of hydraulic parameters in the concentration and spatial distribution of heavy metals in sediments in a two-stage reservoir
No full textThe role of hydraulic parameters in sediment transport and heavy metals concentration still needs scientific research. In this study, GIS techniques, IBER (a 2D hydrodynamic modeling system) and statistical analysis were applied to assess heavy metals concentration, spatial distribution and sources. A total number of 30 surface sediment samples were collected from the Stare Miasto twostage reservoir. Results showed that median values follow order Zn > Pb > Cu > Cr > Ni > Cd, which was characteristic for both parts of the reservoir. The overall calculated median concentrations of Zn, Pb, Cu, Cr, Ni, and Cd were 6.74, 1.66, 1.14, 0.99, 0.8, and 0.04 mg/kg. Analysis of heavy metals concentration shows that higher mean values were observed in the pre-dam part for all of the analyzed heavy metals. The highest risk was observed for Zn, Pb and Cd for all of the analyzed samples. Statistical analysis showed that heavy metals concentration is correlated with the fraction of sediments and distance from the inflow. Spearman’s rank correlation showed that hydraulic parameters affect heavy metals concentration. Critical diameter was negatively correlated with Cu while Froude number and velocity were negatively correlated with Cu and Zn concentrations. Also, it was observed that Cu concentrations in the main zone were positively correlated with specific discharge. Results showed that the two-stage construction of the reservoir has an impact on the limitation of sediments spatial distribution and helps to control pollution related to heavy metals.Postprint (published version
Recovering Einstein’s mature view of gravitation: a dynamical reconstruction grounded in the equivalence principle
No full textThe historical and conceptual foundations of General Relativity are revisited, putting the main focus on the physical meaning of the invariant ¿¿2 , the Equivalence Principle, and the precise interpretation of spacetime geometry. It is argued that Albert Einstein initially sought a dynamical formulation in which ¿¿2 encoded the gravitational effects, without invoking curvature as a physical entity. The now more familiar geometrical interpretation—identifying gravitation with spacetime curvature—gradually emerged through his collaboration with Marcel Grossmann and the adoption of the Ricci tensor in 1915. Anyhow, in his 1920 Leiden lecture, Einstein explicitly reinterpreted spacetime geometry as the state of a physical medium—an “ether” endowed with metrical properties but devoid of mechanical substance—thereby actually rejecting geometry as an independent ontological reality. Building upon this mature view, gravitation is reconstructed from the Weak Equivalence Principle, understood as the exact compensation between inertial and gravitational forces acting on a body under a uniform gravitational field. From this fundamental principle, together with an extension of Fermat’s Principle to massive objects, the invariant ¿¿2 is obtained, first in the static case, where the gravitational potential modifies the flow of proper time. Then, by applying the Lorentz transformation to this static invariant, its general form is derived for the case of matter in motion. The resulting invariant reproduces the relativistic form of Newton’s second law in proper time and coincides with the weak-field limit of General Relativity in the harmonic gauge. This approach restores the operational meaning of Einstein’s theory: spacetime geometry represents dynamical relations between physical measurements, rather than the substance of spacetime itself. By deriving the gravitational modification of the invariant directly from the Weak Equivalence Principle, Fermat Principle and Lorentz invariance, this formulation clarifies the physical origin of the metric structure and resolves long-standing conceptual issues—such as the recurrent hole argument—while recovering all the empirical successes of General Relativity within a coherent and sound Machian framework."This research was founded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe” grant number PID2021-123903NB-I00, by program Unidad de Excelencia María de Maeztu grant number CEX2020-001058-M and by AGAUR grant number project 2021-SGR-00171"Peer ReviewedPostprint (published version
Determinants of long-term SARS-CoV-2 immune responses in asymptomatic-to-moderate COVID-19 patients in sub-Saharan Africa
No full textBackground Immune responses after SARS-CoV-2 infection remain poorly characterized in African populations, despite widespread viral transmission and proportionally lower COVID-19 severity and mortality than in other regions. We aimed to define the determinants and durability of humoral and cellular immunity in sub-Saharan Africa and to identify immune correlates of protection against reinfection. Methods We conducted a 12-month longitudinal immunological study involving 513 adults with asymptomatic or mild-to-moderate COVID-19 enrolled across four sub-Saharan African countries (Ghana, Democratic Republic of Congo, Ethiopia, and Mozambique) during four pandemic waves (2020–2022). We profiled levels of IgA, IgG, and IgM against eight SARS-CoV-2 antigens and neutralizing antibody activity against ancestral and variant strains by Luminex, and antigen-specific T- and B–cell responses by flow cytometry. Immune kinetics, decay, immune escape, and reinfection risk were evaluated alongside the impact of clinical and demographic variables, including prior exposure, epidemic wave, geographic site, treatment allocation, and host factors. Statistical analyses included non-parametric tests (Kruskal–Wallis with Benjamini–Hochberg adjustment), Spearman correlations, logistic regression for reinfection, and mixed-effects models for longitudinal determinants. Results Humoral and cellular immune responses were robust and sustained across participants. Estimated antibody half-lives during the early decay phase exceeded 50 days for IgA and IgG. Higher IgA, IgG, and neutralizing levels were significantly associated with lower odds of reinfection during follow-up. Repurposed COVID-19 treatments showed no measurable impact on immune responses. Prior infection and vaccination were the main determinants of antibody magnitude and persistence, greatly surpassing the effects of age, sex, symptoms, and comorbidities. Antibody levels also varied significantly by epidemic wave and site, higher in later waves and, across sites, generally higher in Ethiopia and lower in DRC. Comorbidities were primarily associated with increased SARS-CoV-2–specific T-cell activation. Strong correlations were observed between binding and neutralizing antibodies, and variant-specific immune escape was confirmed for Beta, Gamma, and Omicron. Conclusions This multi-country study provides a comprehensive characterization of SARS-CoV-2 humoral and cellular immune responses in African cohorts and identifies prior exposure and local epidemiological context as the main determinants of immune magnitude, durability, and protection, outweighing other host factors.Major funding for the ANTICOV and ANTICOV–IMMUNO consortium was provided by the German Federal Ministry of Education and Research (BMBF) through KfW and by the global health agency Unitaid as part of ACT-A. Early support to launch the initiative was provided by the European & Developing Countries Clinical Trials Partnership (EDCTP), under its second program supported by the European Union with additional funding from the Swedish government, and the Starr International Foundation, Switzerland. The serological work was also supported by the Fundació Privada Daniel Bravo Andreu. G.M. was supported by grant no. RYC 2020–029886-I/AEI/10.13039/501100011033, co-funded by the European Social Fund (ESF). We acknowledge support from the grant CEX2023-0001290-S funded by MCIN/AEI/10.13039/501100011033, and support from the Generalitat de Catalunya through the Centres de Recerca de Catalunya (CERCA) Program.Peer Reviewed10 - Reducció de les Desigualtats3 - Salut i BenestarPostprint (published version