
Recent Additions
- Some of the metrics are blocked by yourconsent settings
Thesis Precise determination of the axial contribution to the coherent elastic neutrino-nucleus scattering event rate(Universidad Técnica Federico Santa María, 2026-04-24)Coherent elastic neutrino–nucleus scattering (CEνNS) is a low-energy process within the Standard Model, characterized by an enhanced cross section due to the coherent vector weak charge. Although this dominant contribution has been thoroughly investigated, subleading axial-vector effects are typically considered suppressed and have consequently received limited attention. In this work, we present a systematic determination of the axial contribution to the CEνNS differential cross section within an effective field theory framework, incorporating spin-dependent nuclear structure factors from modern nuclear calculations. We identify target-dependent regimes in which the axial contribution is not strongly suppressed relative to the vector term, challenging the common assumption of its negligibility and providing a basis for precision CEνNS analyzes beyond the coherent approximation. - Some of the metrics are blocked by yourconsent settings
Thesis Efecto de un rejuvenecedor basado en la pirólisis de neumáticos sobre el comportamiento mecánico de mezclas asfálticas frías recicladas con emulsión(Universidad Técnica Federico Santa María, 2026-04-09)La incorporación de pavimento asfáltico recuperado (RAP) en mezclas asfálticas en frío se presenta como una estrategia fundamental para lograr la sustentabilidad en la industria de la construcción, sin embargo, cuenta con numerosos desafíos técnicos dados por su inherente rigidez como consecuencia del envejecimiento del ligante asfáltico, la presente investigación evalúa la factibilidad mecánica de esta tecnología al incorporar un agente rejuvenecedor basado en la pirólisis de neumáticos fuera de uso (LP) con la capacidad de restaurar las propiedades viscoelásticas de cementos asfálticos envejecidos, durante el estudio se evaluaron ensayos de Módulo Dinámico, Número de Flujo y Resistencia a Tracción Indirecta, los cuales son fundamentales para caracterizar el comportamiento que tiene la aplicación de diferentes dosis de RAP (0%, 50% y 100%) y de LP (9%, 15% y 21%), los resultados demostraron que la adición del agente rejuvenecedor logra modificar la reología del ligante envejecido reduciendo la rigidez e incrementando la zona de deformación secundaria, sin embargo, excesivas dosis de LP presentaron desempeños inadecuados producto de un excesivo reblandecimiento del material previsiblemente debido a una sobresaturación de la matriz agregados pétreos-ligante asfáltico-agente rejuvenecedor, en adición el comportamiento de mezclas compuestas en 100% RAP con diferentes dosis de LP presentó resultados altamente variables y poco predecibles, mientras que mezclas asfálticas conformadas por 50% agregados vírgenes (A.V.) y 50% RAP con dosis de LP menores a 15% mostraron un adecuado comportamiento ante daño por humedad, una apropiada resistencia a deformación terciaria y un módulo dinámico similar a las mezclas confeccionadas con 100% A.V., logrando beneficios técnicos al utilizar este tipo de tecnologías. - Some of the metrics are blocked by yourconsent settings
Thesis High-density surface electromyography as a method for estimating intrinsic laryngeal muscle activity(Universidad Técnica Federico Santa María, 2026-03-06)During speech, the activity and coordination of the intrinsic laryngeal muscles are essential for modulation and control, with the cricothyroid (CT) muscle being primarily responsible for regulating the fundamental frequency during phonation. Currently, methods exist for estimating the activity of these muscles; however, their applicability is limited by the invasive nature of intramuscular electromyography (iEMG) and by the spatial superposition inherent to conventional surface recordings. This thesis proposes and evaluates a non-invasive framework for estimating CT muscle activation using high-density surface electromyography (HDsEMG), supported by synchronized intramuscular recordings as a physiological reference. First, an initial experiment was conducted in which a dataset of simultaneous iEMG and acoustic recordings was established in healthy subjects at the Massachusetts General Hospital (MGH). In this experiment, an acoustic-based normalization strategy anchored to the fundamental frequency (f0) and sound pressure level (SPL) was developed to improve inter-session and inter-task consistency of EMG measurements. These data allowed the creation of a database that served as the gold standard for validating the subsequent results. Subsequently, in a second experiment, 13 patients were recorded at the VPLab laboratory of AC3E, acquiring simultaneous iEMG, HDsEMG, and acoustic signals. The HDsEMG and iEMG signals were decomposed into motor unit discharge trains and validated using a two-source approach with concurrent intramuscular decomposition, enabling the identification of motor units detected by both techniques. The spatial characterization of motor units was performed by estimating the center of gravity and inferring depth based on Gaussian surface fitting, which allowed the separation of superficial and deep activation regions within the CT muscle.For each spatial group, cumulative spike trains (CSTs) were computed and compared with intramuscular EMG envelopes using correlation, coherence, and error metrics. Although the CSTs reliably reproduced temporal trends and shared synaptic input with the iEMG activity, amplitude mismatches were observed, particularly during dynamic phonatory tasks. To address this limitation, a subject-specific neural network was implemented to estimate normalized CT activation by combining CST-derived features with acoustic correlates (f0 and SPL). This data-driven approach improved amplitude estimation while preserving physiologically meaningful temporal dynamics. Overall, the results demonstrate that HDsEMG, when combined with spatially informed processing, acoustic normalization, and machine learning, enables the non-invasive estimation of an intrinsic laryngeal muscle activation, providing a foundation for future studies in voice production modeling and potential clinical assessment. - Some of the metrics are blocked by yourconsent settings
Thesis Predicción del desempeño de pila de combustible de hidrógeno verde en la Región de Antofagasta mediante modelos de Aprendizaje Automático(Universidad Técnica Federico Santa María, 2026-03-06)El despliegue de tecnologías de hidrógeno verde en la Región de Antofagasta requiere herramientas predictivas que permitan evaluar el desempeño de bancos de pilas de combustible de membrana de intercambio protónico (PEMFC) en un entorno con alta heterogeneidad climática y topográfica. En este contexto, se desarrolla un pipeline de modelado basado en Aprendizaje Automático para predecir con alta precisión y eficiencia la potencia eléctrica de un banco de tres pilas GenSure E-1100 de la planta piloto móvil de hidrógeno verde de CICITEM, desplegada desde la Cordillera de la Costa hasta la Precordillera y la Depresión Intermedia. La base de datos utilizada está compuesta por 1595 registros de variables ambientales (Tamb, pamb, HR) y eléctricas (I, V, W) recopilados en distintos sitios, conformando un conjunto de tamaño acotado, pero altamente heterogéneo y multimodal, con desbalances en los estratos climático-operacionales y una marcada presencia de outliers, lo que exige esquemas de modelado robustos y coherentes con la física del sistema. El flujo de trabajo integra tres componentes principales: (i) ingeniería de características acoplada a ablación incremental para construir una base de datos enriquecida con descriptores climáticos, temporales, operacionales y de reparto de combustible; (ii) una metodología probabilística de detección de outliers por familias de variables (FASEK5), que combina múltiples detectores en un esquema leaky noisy-OR para generar versiones depuradas de la base preservando su representatividad física y operativa; y (iii) la implementación de un pipeline robusto y reproducible para entrenar y comparar modelos de regresión (RandomForestRegressor, XGBoost, CatBoost, MLPRegressor y SupportVectorRegressor) con validación cruzada estratificada y optimización bayesiana de hiperparámetros. La calidad de los modelos se evalúa mediante un conjunto integrado de métricas de ajuste, sesgo y dispersión de residuos (RMSE, r, R2, CCC, MEC, ME, SDE, Std), complementado con diagramas solares y de Taylor, además de métricas de generalización y ganancia relativa asociada a la depuración de la base de datos (GAPRMSE, ΔRMSE, Gainrel). Los resultados muestran que la depuración moderada de outliers (< 5% de contaminación, < 80 registros) incrementa de forma sistemática la capacidad de generalización de todos los modelos, con reducciones de la diferencia entre el RMSE de entrenamiento y de prueba (GAPRMSE) del orden de 4.6–9.0 W y ganancias relativas del RMSE de prueba en torno a 21–31 %, desplazando sus posiciones en los diagramas solar y de Taylor hacia regiones de mayor calidad. Bajo este marco metodológico, la configuración XGB v0.9950 emerge como el modelo con mejor compromiso entre precisión, robustez frente a outliers y capacidad de generalización, alcanzando un RMSE global ≈ 31 W (≈ 11 % de la desviación estándar experimental, σ ≈ 275.6 W), sesgo prácticamente nulo y valores de r ≈ 0.993 y R2 ≈ 0.987, con coeficientes de eficiencia de modelado (MEC) y de concordancia (CCC) del orden de 0.990. El análisis de importancia de características en los modelos basados en árboles indica que el comportamiento del banco está gobernado principalmente por el régimen de activación, la configuración interna y el reparto de combustible entre pilas y, en segundo lugar, por las condiciones ambientales y el contexto temporal, en concordancia con la influencia sobre el desempeño electroquímico predicha por modelos analíticos. Este trabajo de tesis valida que un enfoque de Aprendizaje Automático integrado con ingeniería de características coherente con la física del sistema PEMFC y el contexto operativo local, junto con una etapa de depuración probabilística, constituye una herramienta robusta para predecir el desempeño de bancos PEMFC en campo. En consecuencia, el pipeline propuesto se configura como un marco reproducible extrapolable a otros bancos y contextos climáticos y sienta las bases para futuros trabajos sobre control, optimización e interpretabilidad de sistemas PEMFC bajo condiciones reales de operación. - Some of the metrics are blocked by yourconsent settings
Thesis Estudio numérico del comportamiento de dispositivos U-shaped en acero y aleaciones con memoria de forma (SMA) a base de Cu para sistemas multidireccionales de disipación de energía(Universidad Técnica Federico Santa María, 2026-03-20)In this work, the mechanical behavior of U-Shaped Crawler-type devices used as seismic energy dissipation elements is analyzed, with particular emphasis on their response under out-of-plane loading (90◦). The methodology integrates analytical formulation, finite element simulations implemented in ANSYS, and statistical learning techniques based on Gaussian Process Regression (GPR), considering multiple geometric configurations and two materials: structural steel A36 and a Cu–Al–Be shape memory alloy with nominal composition Cu = 88%, Al = 12%, Be = 0.48% (CAB0.48–A(-24)). The numerical model was developed based on the three-dimensional geometry of the U-Shaped Crawler-type device, employing SOLID187 elements. Using principles of energy and static equilibrium, analytical expressions were derived to estimate the initial stiffness and yield force, which were subsequently contrasted with 144 numerical simulations defined through a factorial design in the parameters L, H, B, and t. From the analysis of absolute and relative errors between analytical predictions and numerical results, functional geometric ranges (1.5 ≤ L/t ≤ 5; 2 ≤ H/B ≤ 4) were identified, within which the analytical model exhibits bounded error levels and consistent agreement. To enhance the predictive capability of the analytical model developed for the U-Shaped Crawler-type element, Gaussian Process Regression (GPR) models were trained, showing outstanding performance in predicting the initial stiffness (R2 = 1.000, RMSE < 0.01 kN/mm) and an adequate fit for the estimation of the yield force. Independently, the behavior of a multidirectional device composed of eight orthogonally arranged U-Shaped elements was evaluated through finite element method (FEM)-based numerical simulations under cyclic loading in the 90◦ direction. The mixed configuration (steel + CuAlBe) exhibited the best balance among stiffness (9.9 N/mm), energy dissipation (0.27 J/cycle), and recovery capacity (IRD ≈ 0.94), indicating promising performance for multidirectional applications. Overall, the results consolidate a methodology that integrates analytical formulation, numerical simulation, and statistical learning to estimate key structural parameters based on the geometry of the U-Shaped Crawler-type element, defined by its characteristic dimensions (L, H, B, and t). Within this framework, GPR is positioned as a complementary tool for the efficient exploration of the design space, reducing the need for a large number of additional numerical simulations.
Most viewed
- Some of the metrics are blocked by yourconsent settings
Thesis Risk-Aware Portfolio Optimization via Reinforcement Learning with Expected Shortfall(Universidad Técnica Federico Santa María, 2025-11-24)En esta tesis se realiza el desarrollo de un modelo de Reinforcement Learning para optimización secuencial de portafolios, incorporando el riesgo mediante el Expected Shortfall. El modelo aprende políticas de decisiones sobre activos de un portafolio (comprar, vender o mantener) usando información de retornos de distintos tipos de activos (acciones, ETFs y criptomonedas) para distintas líneas temporales, junto con un umbral de decisión compuesta que se adapta de forma dinámica según la volatilidad de los retornos. Basándose en prueba y error, obteniendo como recompensa una métrica de retorno ajustado por riesgo, definida por el retorno marginal obtenido de un conjunto de movimientos respecto a una inversión libre de riesgo, ajustada por una métrica de riesgo explícita.El objetivo principal de la tesis es evaluar si un enfoque de Reinforcement Learning logra detectar patrones dinámicos de mercado y lograr generalizarlos para escenarios fuera de muestra. Comparándolo con otros modelos tanto financieros estadísticos, como modelos de machine learning supervisados y no supervisados. Los resultados muestran que, para enfoques más conservadores sobre el umbral, el modelo de Reinforcement Learning tiene un menor gap entre entrenamiento validación y testeo, siendo sus resultados más permanentes entre fases. Los enfoques más de rentabilidad pura, poseen mayor volatilidad entre fases de validación y testeo, teniendo un mayor riesgo que sus competidores y resultados similares. - Some of the metrics are blocked by yourconsent settings
Thesis Producción y caracterización de cuerpos de inclusión recombinantes para ser usados como antígenos en un prototipo de vacuna contra Piscirikettsia salmonis(Universidad Técnica Federico Santa María, 2025-10-20)Piscirickettsia salmonis is the etiological agent of Piscirickettsiosis, one of the most severe and detrimental diseases affecting the Chilean salmon farming industry. Currently, there are 26 vaccines with provisional registration available for prophylactic use in salmonids. However, none have proven effective in fully controlling epidemic outbreaks under field conditions. Several researchers have identified two predominant genogroups along the Chilean coast, designated as LF-89 and EM-90, which exhibit a variable spatiotemporal distribution in southern Chilean salmon farms in recent years. Recent evidence suggests that the biological mechanisms of both genogroups may interact synergistically during infection, indicating the potential for co-infections in contemporary farms. This scenario further complicates vaccine efficacy, as many are designed to target only a predominant genogroup. Thus, the need arises to identify alternative prophylactic methods to ensure and contribute to the sustainability of the Chilean salmon farming industry. A novel prophylactic approach proposed in this thesis involves the use of recombinant nanoproteins, also known as inclusion bodies, with antigenic functions against Piscirickettsiosis. These inclusion bodies are aggregates of soluble and insoluble proteins with amyloid characteristics, eliminating the need for encapsulation. They are scalable, lyophilizable, and highly stable under adverse temperature and pH conditions. This thesis aims to evaluate the ability of certain antigenic protein inclusion bodies from P. salmonis to induce an adaptive immune response in Atlantic salmon (Salmo salar) and their potential use as immunogens in a vaccine prototype. To achieve this objective, three antigenic sequences with chimeric characteristics from the LF-89 and EM-90 genogroups of P. salmonis were designed and produced as recombinant inclusion bodies in E. coli BL21(DE3). The prototypes were designated for commercial protection purposes as SkipZ, PulseJ, and HopQ. These prototypes were quantified and characterized using techniques such as Western blot. Their uptake and immunostimulatory activity were evaluated in salmon macrophages through in vitro assays. Subsequently, juvenile salmon were vaccinated, and the effects of the SkipZ and HopQ prototypes were analyzed in terms of gene expression (qPCR) and immunoglobulin production (ELISA). The three prototypes achieved production efficiencies of 57 mg/L for PulseJ, 40 mg/L for SkipZ, and 26,8 mg/L for HopQ. These prototypes were internalized by salmon macrophages (RTS-11) with uptake efficiencies of 5% for SkipZ, 26% for PulseJ, and 54% for HopQ. In vitro assays showed activation at both transcriptional and translational levels of molecules associated with antigen presentation and pro-inflammatory markers, with notable activation capacities observed for HopQ and SkipZ at 20 µg/mL and 5 µg/mL, respectively. In vivo assays revealed that both vaccine prototypes, administered at doses of 0.5 mg/kg of fish for SkipZ and 2 mg/kg of fish for HopQ, robustly activated the Th1 response, with a marked increase in IFN-γ and IL-12 expression. This strong upregulation might suppress CD8+ lymphocyte differentiation, potentially reducing the effectiveness of the cellular immune response in fully eliminating the pathogen. However, this effect may contribute to maintaining homeostasis in fish during the exacerbated inflammation induced by the vaccines. Moreover, no significant change was observed in the Th2 response, suggesting that the humoral response was not predominantly activated. In conclusion, the vaccine prototypes based on antigenic sequences of Piscirickettsia salmonis and produced as chimeric inclusion bodies demonstrated potent immunogenicity, capable of inducing a robust adaptive immune response in Atlantic salmon. The activation of the Th1 response by HopQ and SkipZ was particularly noteworthy. However, further studies are needed to evaluate their protective effects in challenge trials with the pathogen, including both genogroups, and to determine their long-term effectiveness. - Some of the metrics are blocked by yourconsent settings
Thesis Revisión y Actualización del Método de Reconciliación del convertidor Teniente en Fundición Hernán Videla Lira ENAMI(2016)The main objective of this study was to propose improvements to current accounting metallurgical Balancer in Hernán Videla Lira Smelter, focusing on balances carried on the main casting equipment, Teniente Converter. These improvements are based on phenomenology of concentrate smelting by Teniente Converter, which are not considered in the current balancer.Because the current balancer is based on a purely mathematical model, the phenomenologically unlikely results, although not common, are possible. Hence adding phenomenological restrictions on the balancer would seek to eliminate these results. In this study the use of two relations are proposed as a basis for the phenomenological restrictions; the relation between white metal copper grade and slag copper grade, and the relation between white metal copper grade and the oxygen coefficient. In addition, it is proposed in parallel with these restrictions, add to the balance, the gas streams that go in and out the Teniente Converter, which are currently not considered.Regarding gas streams, these must be added because not considering them create a 1 monthly ton imbalance for every 5 monthly tons of molten concentrate, and considering that on average about 25.000 tons of concentrate melt to month, the product imbalance of not consider gas streams is 5.000 tons.Regarding the white metal copper grade and slag copper grade, this relation is primarily used when the copper grade of the slag is too high, while the relation between white metal copper grade and the oxygen coefficient is used mainly in cases where the copper grade of the white metal is too low.When analyzing the results it was observed that different and possibly better results are obtained when using phenomenological relations than with the current model. In addition, it was analyzed how restrictions influence on the results when the initial data were significantly unbalanced regarding weights and / or with respect to the fine copper, in scenarios without gases, with gases and only considering exhaust gases, where it was concluded that independent of the scenario, the balancer always tends to modify the output currents to equalize to the input, due to the effect of the quality factors of every current.It was also seen that in general, arsenic emissions increase or decrease depending on whether the initial imbalance is a deficiency or excess respectively of the input currents for both the fine and for the weights, where they may vary by up to 198% in some cases. - Some of the metrics are blocked by yourconsent settings
Thesis Estudio de adhesión de granos de arena sobre vidrio nanoestructurado a través de microscopía de fuerza átomica(Universidad Técnica Federico Santa María, 2025-06-17)Los vidrios nanoestructurados han demostrado ser altamente efectivos para aplicaciones en sistemas ópticos y optoelectrónicos, ya que permiten optimizar la transmitancia óptica y conferir propiedades autolimpiantes, especialmente útiles en contextos como la generación de energía solar. Este trabajo se centró en el estudio de la adhesión de partículas de arena sobre vidrios con distintas morfologías superficiales, utilizando espectroscopía de fuerza mediante Microscopía de Fuerza Atómica (AFM). Para esto se emplearon sondas coloidales fabricadas en el laboratorio, utilizando granos de arena reales recolectados en la Plataforma Solar del Desierto de Atacama (PSDA), los cuales fueron previamente separados en fracciones solubles e insolubles. Las sondas fueron caracterizadas mediante microscopía electrónica de barrido (SEM) y espectroscopía de rayos X (EDX), confirmando la composición mineralógica reportada en estudios previos. Posteriormente, se realizaron mediciones de fuerza de adhesión, usando la técnica de sonda coloidal, sobre tres tipos de superficie: vidrio liso, vidrio tratado químicamente con KOH y vidrio nanoestructurado, todos fabricados usando la técnica propuesta por Bravo et al. Los resultados muestran una disminución significativa de la fuerza de adhesión para partículas solubles sobre superficies nanoestructuradas en comparación con vidrio liso, lo cual se atribuye a la reducción del área efectiva de contacto provocada por la rugosidad. En cambio, al emplear partículas solubles e insolubles, se observó un aumento de la fuerza de adhesión sobre vidrio tratado con KOH, a pesar de su mayor rugosidad, lo cual sugiere una intensificación de fuerzas capilares asociadas al tratamiento químico. Este estudio demuestra que la técnica de sonda coloidal con partículas reales permite caracterizar de manera precisa la interacción entre contaminantes y superficies funcionalizadas. Asimismo, valida el método de nanoestructuración propuesto como una estrategia viable para mitigar el fenómeno de soiling. - Some of the metrics are blocked by yourconsent settings
Thesis Pion production by diffraction of W bosons in the dipole model(2011-06)In this work the scattering of neutrinos off heavy nuclei by exchange of W bosons in the color dipole model is studied, for small boson virtualities and energies of the order of some GeV’s. The first objective of this work is to analyze the black disc limit and to compare the results of the dipole model using perturbative light front wave functions with the ones of the meson model, which are vanishing because of lepton current conservation. Since both results differ, a more complete analysis is carried out, based on Glauber’s approach. At high energies, it is known that the color dipole model is well suited to study diffractive processes on nuclei, but at smaller energies the dipole lifetime is not large enough for the dipole to travel through the nucleus with a frozen transverse size. The short-lived dipoles are identified, and explicit calculations are carried out to determine if they have a large contribution to the final cross section. Although the model presents serious problems because of these short-lived asymmetric pairs, it provides a lower bound for the total cross section of pion production and it gives some additional hints that could be also of interest in more sophisticated models.
