Doctoral School

Reading date: 02/04/2025

• PÉREZ I GONZALO, RAÜL: End-to-end learning for wind turbine blades: from imagery data to defect repair recommendations
  
  Author: PÉREZ I GONZALO, RAÜL
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN AUTOMATIC CONTROL, ROBOTICS AND VISION
  Department: Institute of Robotics and Industrial Informatics (IRI)
  Mode: Normal
  Deposit date: 26/02/2025
  Reading date: 02/04/2025
  Reading time: 11:00
  Reading place: Sala d'Actes de la Facultat de Matemàtiques i Estadística (FME) de la Universitat Politècnica de Catalunya, C/de Pau Gargallo, 14, 08028 Barcelona
  Thesis director: AGUDO MARTÍNEZ, ANTONIO
  Thesis abstract: The European Union's (EU) reliance on external energy sources underscores the urgent need for energy security and affordability, driving the transition to renewable energy with wind power as a key renewable solution. However, wind turbine operation and maintenance still account for 30% of energy production costs, due to their prolonged exposure to harsh environmental conditions. Timely defect detection and repair are critical, as turbines must often be halted during visual inspections and repairs. Streamlining the process from inspection to decision-making is essential to reduce downtime and operational costs.This thesis presents a comprehensive end-to-end blade assessment system designed to determine defect severity, quantify their impact on energy production, and deliver actionable repair recommendations. By enabling wind turbine owners to act proactively, this system helps minimize operational costs. The framework emphasizes efficient image transmission that preserves quality, followed by the generation of detailed blade assessments to establish a consistent and effective repair strategy.To this end, this project proposes first segmenting images to isolate blade regions, simplifying subsequent tasks through algorithms tailored for imagery acquired under diverse conditions. These include a Blade U-Net model, which introduces dense conditional-random-field regularization to enhance segmentation accuracy, and advanced post-processing involving iterative refinement through hole-filling and noise reduction via an unsupervised random forest. Two deep discriminant analysis frameworks integrate class separability and probabilistic modeling into robust non-linear architectures to derive precise defect boundaries, handle complex textures, and improve generalization across varied inspection data. Additional contributions include a modular region-growing classifier for efficient segmentation in data-scarce conditions and diffusion-based models with dual-space augmentation to improve generalization and robustness, leading to substantial superior performance than competing techniques. Together, these segmentation methods form the foundation for automated defect detection and diagnostics.In the second part, to address the challenge of handling large volumes of high-resolution inspection data, this work also presents a novel region-of-interest (ROI) image compression framework. Traditional methods often compromise critical defect information. The proposed framework leverages segmentation outputs to ensure high-fidelity compression in blade regions, employing lossless or high-quality lossy techniques while aggressively compressing non-relevant areas. Key innovations include multi-layer nested latent variable models for lossy coding and parallelized bits-back coding optimized for industrial-scale applications. These advancements achieve state-of-the-art performance while significantly reducing computational costs. By coupling compression with our proposed multi-task defect detection model, this approach supports timely and accurate diagnostics, ensuring minimal disruption to turbine operations.In summary, this thesis contributes a hierarchy of low-level to high-level algorithms designed to streamline wind turbine maintenance processes. The combination of advanced segmentation and compression enables a fully automated pipeline for blade defect assessment, encompassing defect localization, classification, and repair prioritization, directly improving energy efficiency by reducing downtime, optimizing maintenance schedules, and minimizing repair costs.
  

Reading date: 04/04/2025

• AL HANAINEH, WAEL HASAN AHMAD: Designing and Development of Secure Protection Strategies for Distribution Network Integrated with Distributed Energy Resources
  
  Author: AL HANAINEH, WAEL HASAN AHMAD
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ELECTRICAL ENGINEERING
  Department: Department of Electrical Engineering (DEE)
  Mode: Article-based thesis
  Deposit date: 10/03/2025
  Reading date: 04/04/2025
  Reading time: pending
  Reading place: pending
  Thesis director: MATAS ALCALA, JOSE | GUERRERO ZAPATA, JOSEP MARIA
  Thesis abstract: Global electricity generation increasingly incorporates Distributed Generation (DG) resources, such as solar and wind, into distribution systems (DS), offering benefits like improved reliability, power quality, rapid integration, and reduced payback periods, while lowering greenhouse gas emissions. However, their integration presents challenges, including overvoltage, voltage fluctuations, and imbalances caused by improper synchronization with the grid. DGs alter short-circuit currents, necessitating updates to protection relay settings. As DG penetration rises, distribution networks become more complex, requiring advanced protection systems to handle bidirectional power flows, which challenge existing schemes. Inverter-based DGs, such as solar and wind, generate lower fault currents due to inverter power electronics, diminishing the effectiveness of traditional fault detection methods, leading to potential protection blinding or false tripping. These challenges highlight the need for precise fault detection, accurate localization, and rapid protective responses. Disconnecting DGs during faults is increasingly undesirable, requiring innovative protection schemes to minimize unnecessary disconnections and address limitations like fault resistance, pre-fault load conditions, and noise interference. Traditional fault location techniques, often computationally intensive, struggle with accuracy, prolonging restoration times and increasing downtime, further emphasizing the need for advanced fault protection systems. Total Harmonic Distortion (THD) analysis has proven effective for fault detection in systems with complex harmonic profiles caused by DG integration. Faults induce increased harmonic distortion, making THD monitoring a valuable indicator. Despite its promise, protection systems for grids with high DG penetration, especially those using inverter-based DGs, are underexplored, and existing protection algorithms rarely incorporate THD. To address this, three novel protection systems utilizing grid voltage harmonic content for fault detection and localization in medium-voltage (MV) DS are proposed. The first system combines THD measurements with voltage amplitude and zero-sequence components using a finite state machine (FSM)-based algorithm. It focuses on third harmonic (triple-n) components, unique to inverter neutral points and unaffected by other grid harmonics. Fault-induced voltage dips excite harmonic components, amplifying THD, making it an effective fault indicator. THD is calculated using the Multiple Second Order Generalized Integrator (MSOGI) method. However, this system relies on communication channels, which could fail, limiting its robustness. To mitigate this, a two-layered protection system is introduced. The first layer employs the SOGI-FLL grid monitoring technique, optimizing computational efficiency by reducing the number of required SOGIs while maintaining accurate THD calculations. Fault detection is achieved by filtering the THD signal and comparing pre-fault and fault-time averages, with significant deviations indicating faults. The second layer implements a communication-less fault localization algorithm based on positive and negative voltage sequence components to determine fault symmetry. This approach enables each protection device (PD) to operate independently, ensuring reliable fault localization even without communication, albeit with slightly slower detection times compared to communication-based methods. To enhance overall reliability, especially during communication failures, a third system, priority system, is proposed. It integrates the two-layered protection, with the first layer as the primary fault detection and communication-based trip signal initiator. If communication fails, the second layer provides backup protection by analyzing voltage sequence components locally. The effectiveness of these systems is validated against different protection method under various conditions.
  

• MORENO MARTÍN, SIRO: Collocation methods for the synthesis of graceful robot motions
  
  Author: MORENO MARTÍN, SIRO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN MECHANICAL, FLUIDS AND AEROSPACE ENGINEERING
  Department: Department of Mechanical Engineering (EM)
  Mode: Normal
  Deposit date: 10/03/2025
  Reading date: 04/04/2025
  Reading time: pending
  Reading place: pending
  Thesis director: CELAYA LLOVER, ENRIC | ROS GIRALT, LLUIS
  Thesis abstract: Graceful motion can be loosely defined as the one we observe in natural movements executed by animals and humans, which are characterized by being agile, efficient, and fluid. The generation of graceful robot motions is typically sought through the minimization of cost functions involving not only path length, but also aspects related to smoothness, like the time derivative of acceleration, called jerk, or that of the controls. A widely used approach to compute optimal trajectories is through direct collocation, a technique that converts the continuous-time optimal control problem into a finite-dimensional NLP problem. Collocation proceeds by discretizing the trajectory using so-called collocation points, and imposing the dynamics constraints at such points. The formulation of most collocation methods, however, assumes that the system is governed by a first order ODE, whereas robotic systems are typically described by second or higher order ODEs. As a result, the usual practice is to initially convert those ODEs into first order form via introducing new variables, and adding new equations that link these variables with their integral counterparts. An often overlooked effect of this transformation is that it generates inconsistencies between the trajectory of each variable and that of its time derivative. This is so because a collocation method only imposes the differential relationships at the collocation points, but not in between such points. A closer examination of this effect reveals that the dynamic equations, which should be satisfied in the collocation points, are actually violated in them, despite apparently having been enforced. This thesis introduces new collocation methods designed to overcome these problems. Specifically, we develop improved versions of the most popular piecewise and pseudospectral collocation schemes, including the trapezoidal and Hermite-Simpson methods, as well as the LG, LGR, and LGL methods. The new algorithms are able to treat differential equations of arbitrary order M ≥ 1 without having to convert them into first order form. In all of them, the trajectory obtained for each variable coincides exactly with the time derivative of its corresponding integral variable, and the dynamic constraints are satisfied accurately at the collocation points. These properties allow a drastic reduction of the dynamics error of the obtained trajectories in many cases, meaning that the governing equations are better respected along the continuous time horizon of the problem. Our methods also provide trajectories that are smoother than those of conventional ones, and easily treat variables such as jerk or the time derivative of the controls in the cost function. An hp adaptive refinement algorithm is also proposed to combine the benefits of our piecewise and pseudospectral methods so as to speed up convergence to the solutions.
  

• UGRINOVIC KEHDY, NICOLAS: Modeling and Reconstruction of 3D Humans under Context
  
  Author: UGRINOVIC KEHDY, NICOLAS
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ARTIFICIAL INTELLIGENCE
  Department: Department of Computer Science (CS)
  Mode: Normal
  Deposit date: 05/02/2025
  Reading date: 04/04/2025
  Reading time: 15:00
  Reading place: Sala de Juntes de la FME
  Thesis director: SANFELIU CORTES, ALBERTO | MORENO NOGUER, FRANCESC D'ASSIS
  Thesis abstract: The study of human's and their behavior through the analysis of images and videos has long been a central topic in Computer Vision. The reconstruction and modeling of human behavior have garnered increasing attention, due to their potential applications in virtual environments, including AR/VR, sports, fashion, and the film industry. Despite this growing interest, accurately capturing and generating the 3D pose and motion of humans remains an important challenge, primarily due to the vast diversity of human movements and the inherent complexity of the human body. Furthermore, the ability to capture and replicate subtle human interactions---such as a hug---that are intuitively understood by humans continues to be a significant obstacle for machines. This complexity arises from the need for a deep understanding of the physical world, its constraints, and the nuanced ways in which humans interact with it.This thesis presents the development of several methodologies for reconstructing and modeling various aspects of humans in 3D, including detailed shape, pose, and motion, mainly from RGB images. A key emphasis is placed on capturing or incorporating contextual information as additional information. First, we introduce a method for modeling the detailed body shape of individuals, which includes elements such as clothing across a wide range of poses. Subsequently, the focus shifts to the simultaneous pose estimation of multiple individuals, wherein scene constraints are employed to enhance the accuracy of these estimations. This approach addresses the fundamental challenges of depth and scale ambiguity inherent in 3D reconstruction. The work is then extended into the temporal domain, to reconstruct interacting individuals, particularly in scenarios involving close interactions. A significant challenge under such situations is the lack of compliance with physical laws, such as body collisions. To address this, we integrate a fully-featured physics simulator within a motion estimation pipeline to account for these physical inconsistencies. Lastly, we propose a method capable of generating human motion that interacts with a virtual environment. All proposed methods have undergone extensive evaluation.In summary, this thesis introduces a suite of tools for the modeling and reconstruction of 3D humans, advancing the field towards more accurate capture and recreation of realistic behavior for virtual humans, with a particular emphasis on their interactions with its surrounding environment.
  

• ZIVANIC, MILICA: Cold plasma-treated hydrogels for multimodal cancer therapy
  
  Author: ZIVANIC, MILICA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN BIOMEDICAL ENGINEERING
  Department: Department of Materials Science and Engineering (CEM)
  Mode: Change of supervisor
  Deposit date: 10/03/2025
  Reading date: 04/04/2025
  Reading time: 12:00
  Reading place: Aula A1.06, Edifici, A, Av. d'Eduard Maristany, 16, 08019 Barcelona
  Thesis director:
  Thesis abstract: Cold atmospheric plasma (hereon just plasma) is a weakly ionized gas that gained attention as a cost-efficient and well-tolerated cancer treatment that selectively targets the altered redox metabolism of malignant cells. The short penetration depth of direct plasma treatment limits its clinical applications to surface targets. Plasma-treated hydrogels (PTHs) emerge as vehicles for local delivery of therapeutic plasma-derived reactive oxygen and nitrogen species (RONS) to internal targets. To prepare a PTH, an aqueous solution containing low concentrations of polymers is exposed to plasma to enrich it with RONS and is then crosslinked into the three-dimensional hydrogel network entrapping RONS inside. Once in contact with the target, RONS can diffuse from PTH and, above a cell-specific threshold, cause irreversible damage and death to cancer cells. Importantly, PTHs could broaden the clinical application of plasma not only by acting as RONS vehicles but also by being a versatile physicochemical platform that can incorporate different bioactive polymers or drugs for combined therapeutic effects, as explored for the first time in this Thesis.This Thesis proposes and follows an iterative workflow cycle for the development and characterization of PTHs. Here, alginate was chosen as a biopolymer for the preparation of PTHs, due to its biocompatibility, relevance, and versatility in biomedical research, as well as the ability to crosslink under mild conditions. In the first place, an optimized protocol for the preparation of alginate-based PTHs was identified, in order to ensure high retention of therapeutic RONS during the crosslinking process and obtain an injectable, shear-thinning formulation useful for minimally-invasive delivery and shape-adaptability of the PTH. Before this Thesis, biological characterization of PTHs was limited to cancer cytotoxicity reports. Here, the ability of a PTH to induce immunogenic cell death was demonstrated for the first time. As a result, PTH-treated osteosarcoma cells were increasingly phagocytized when co-cultured with immature dendritic cells derived from human monocytes isolated from healthy blood donors. Following the initial physicochemical and biological characterization, the feasibility and efficacy of incorporating a secondary therapeutic modality to the PTH were investigated. For this, a bioactive polymer or a chemotherapeutic drug was introduced into the alginate PTH formulation to achieve biological effects beyond or in synergy with plasma-derived RONS. Importantly, these effects were studied in a relevant model: an in ovo cancer model, where three-dimensional and vascularized tumors were grown on the membrane of a fertilized chicken egg (in ovo). This enabled the assessment of cancer cells in an environment more similar to a native, clinical one. In ovo tumor models emerge as cost- and time-effective models and can help replace, reduce, and refine in vivo experiments in preclinical research. In contrast to mono-therapy with PTH or drug alone, which showed no effect in ovo, a single administration of PTH-drug co-therapy could diminish osteosarcoma tumor weight and the expression of a protein linked to treatment resistance.Altogether, the work presented in this PhD Thesis helped characterize and establish PTHs within the plasma community as a novel modality that can broaden the clinical application of plasma. Furthermore, it provided a proof of concept that PTHs can be used as versatile dual platforms for multimodal cancer management.