https://repositorio.ufu.br/handle/123456789/5465 2026-05-26T02:07:31Z https://repositorio.ufu.br/handle/123456789/48667 Title: Ai-driven lightweight motor assessment embedded in an exergame for post-stroke rehabilitation Abstract: Introduction: Stroke is a leading cause of long-term disability worldwide and frequently results in persistent upper-limb motor impairments. Accurate motor assessment is essential for rehabilitation planning and monitoring recovery. This assessment is traditionally performed using standardized clinical scales, such as the Fugl-Meyer Assessment (FMA), which require trained healthcare professionals, can be time- and resource-intensive, and involve subjectivity, as scores may vary across evaluators. Moreover, they cannot be easily administered remotely, and repeated in-person evaluations can be difficult for individuals with mobility limitations. Faster and more objective assessment approaches could enable more frequent evaluations and improve the tracking of motor recovery. Although recent advances in digital health, exergames, and Artificial Intelligence (AI) have enabled alternative solutions, many existing systems rely on specialized motion capture sensors, time-consuming software, or non-interpretable AI models, limiting accessibility, scalability, and clinical adoption. Objectives: This thesis aims to develop and evaluate a sensor-free, lightweight, and interpretable approach for post-stroke upper-limb motor assessment embedded into a rehabilitation exergame. The proposed system simultaneously provides therapeutic exercises and automatically estimates upper-limb motor performance during gameplay using only a standard camera, enabling supervised remote use. Specifically, the objectives are to (i) review existing automated motor assessment methods and their relationships with kinematic game parameters, and (ii) propose and preliminarily validate a low-cost exergame framework capable of estimating clinical motor impairment using simple and transparent kinematic features. Materials and Methods: A systematic review was conducted to analyze current technologies and computational approaches for automated upper-limb motor assessment after stroke, identifying the most used game parameters. Based on these findings, an experimental study was developed using a Unity-based rehabilitation exergame controlled by arm movements captured with a standard camera and processed using the MediaPipe framework. Sixteen kinematic and spatiotemporal features were extracted from two-dimensional hand and arm trajectories during gameplay. Twelve individuals with chronic stroke (24 upper limbs) participated, with bilateral FMA scores used as the clinical reference. Correlation analyses, exhaustive feature selection, and multiple linear regression modeling were performed to estimate FMA scores, with exploratory comparisons to alternative machine learning models. Results: The systematic review indicated that most existing approaches depend on external sensors or computationally complex methods, often limiting interpretability and real-world applicability. In the experimental study, some gameplay-derived and clinically interpretable features, including average hand aperture, and spatial exploration area, were significantly correlated with FMA scores. A lightweight multiple linear regression model demonstrated strong predictive performance for affected limbs (Spearman ρ = 0.92, R² = 0.89, RMSE = 4.42) and accurately stratified motor impairment severity, achieving accuracies between 86% and 93%. More complex machine learning models did not outperform the interpretable regression approach. Conclusion: This thesis demonstrates that sensor-free, low-cost, interpretable and lightweight computation models for post-stroke motor assessment can be embedded into rehabilitation exergames using kinematic features derived from a standard camera, and be potentially valid for motor function assessment. Further tests with more games and populations are needed to generalize the result. Integrating assessment into gameplay reduces clinical workload, enables high-frequency monitoring, and improves accessibility for telerehabilitation. 2026-03-02T00:00:00Z https://repositorio.ufu.br/handle/123456789/48575 Title: Expressões analíticas aproximadas em forma fechada para a probabilidade média de erro de símbolo da modulação QAM cruzada nos canais de desvanecimento η–μ e κ–μ Abstract: Introduction: Modern communication systems, such as 5G and 6G, require high reliability and low latency, but are heavily affected by channel fading, which compromises signal quality in multipath environments. Although traditional models, such as Rayleigh, Rician, and Nakagami-m, are widely used, they present limitations by assuming specific propagation conditions and failing to accurately represent the diversity of real-world scenarios, especially at millimeter-wave frequencies. To overcome these limitations, generalized models such as the η–μ and κ–μ have been proposed, offering greater flexibility by accounting for multiple dominant components and unbalanced scattering. However, the estimation of error probabilities associated with these models is essential for their practical implementation. Objectives: This work presents the derivation of approximate closed-form analytical expressions for the Average Symbol Error Probability (SEP) and investigates their performance for M-QAM modulation under generalized fading models. Materials and Methods: The adequacy of the derived expressions was analyzed considering standardized millimeter-wave frequency bands of 26, 28, 39, 55, 60, and 65 GHz. These analyses were based on channel modeling results obtained from an extensive measurement campaign in indoor environments under both Line-of-Sight (LoS) and non-Line-of-Sight (nLoS) conditions, as well as varying distances between transmitting and receiving antennas (Tx and Rx) with vertical and horizontal polarizations (VV and HV). Results: The expressions were validated through simulations performed using the MATLAB software, demonstrating strong agreement between the theoretical curves and the simulated results. The analyses revealed that line-of-sight scenarios exhibit significantly superior performance compared to nLoS cases. Furthermore, it was observed that increasing the frequency does not necessarily lead to worse performance, and that higher modulation orders require higher Eb/N0 values to achieve low error rates, highlighting the importance of employing robust coding techniques. The results contribute to a better understanding of the potential and challenges of M-QAM modulation in realistic 5G millimeter-wave communication scenarios. 2026-03-03T00:00:00Z https://repositorio.ufu.br/handle/123456789/48569 Title: Contribuições para o processo da atribuição de responsabilidade das variações de tensão de curta duração baseado nos fatores de desequilíbrio Abstract: Short-Duration Voltage Variation (SDVVs) represent significant deviations in the amplitude of the rms voltage value over a time interval shorter than three minutes. Despite their short duration, such phenomena exert substantial impacts on the quality, reliability, and safety of electrical power systems. Indeed, research studies indicate that the majority of complaints from generation, transmission, distribution agents, as well as end users, are predominantly associated with this type of disturbance. In response, the Brazilian Electricity Regulatory Agency (ANEEL), through the most recent revision of its Distribution Procedures (PRODIST), has introduced preliminary guidelines to regulate these disturbances and ensure adequate power quality levels. Within this context, the issue of identifying the agent responsible for such phenomena emerges as a motivating theme of this doctoral research. In this regard, although several publications addressing the determination of the responsible agent can be found in the literature, there is, to date, no methodology grounded on feasible procedures that enables straightforward field application. Aiming to fill this gap, a previous master’s dissertation developed within the UFU graduate program established the foundational elements of a related procedure. This approach was based on the mechanism of transferring voltage unbalance indicators through the boundary transformer between two agents, by comparing voltage unbalance factor measurements on both sides of the transformer during the occurrence of an SDVV. Notwithstanding the promising perspectives obtained, the methodology was investigated only at an embryonic stage, which motivated the continuation and expansion of the studies through the present doctoral research. Therefore, this work introduces methodological enhancements consisting of: the establishment of conceptual and mathematical foundations to support the proposed process; the extension of its applicability to topological configurations more representative of real-world power systems; and, finally, the experimental validation of the methodology. These advances, which constitute the focus of this research, enable the assessment of the reliability, robustness, and practical applicability of the proposed analytical method. 2026-02-23T00:00:00Z https://repositorio.ufu.br/handle/123456789/48537 Title: Análise Multicritério de Viabilidade de uma Fonte Emergencial Híbrida para Suprimento de Sistema de Serviços Auxiliares em Subestações de Transmissão Abstract: This thesis proposes, evaluates, and implements a solution for the emergency power supply of the auxiliary services alternating current system of high-voltage substations, aiming to enhance operational reliability and reduce regulatory impacts associated with unavailability, particularly in remote substations without internal power sources. The study is grounded in a systematic literature review, which reveals the predominance of arrangements based on diesel generator sets and the lack of integrated approaches that reconcile economic performance, reliability, and environmental criteria. Based on this assessment, a replicable techno-economic and multicriteria evaluation methodology is developed to compare four supply alternatives: a hybrid system combining battery energy storage and photovoltaic generation, an owned diesel generator set, a rented diesel generator set, and an external feeder. The economic analysis employs discounted cash flow methods (NPV, IRR, and Payback), incorporating variable penalties derived from the Annual Allowed Revenue, while the multicriteria assessment applies the TOPSIS method, integrating economic, technical, and environmental criteria. The methodology is validated through a real-world case study conducted at a Brazilian power system substation. The results indicate that the proposed hybrid solution is technically feasible and economically attractive in the medium and long term, standing out for its operational robustness, reduced dependence on fossil fuels, and superior environmental performance, without loss of economic competitiveness, while also presenting potential for scalability and replication. This thesis consolidates an original contribution by demonstrating the technical, economic, and regulatory feasibility of an implementable solution and by structuring a decision-support methodology aligned with the requirements of the Brazilian power sector. 2026-03-06T00:00:00Z