https://ri.ufs.br/jspui/handle/riufs/2442 Wed, 23 Sep 2026 15:07:48 GMT 2026-09-23T15:07:48Z http://ri.ufs.br:80/retrieve/29fb2f39-55c2-4e44-8f88-7fc43105be7a/engeeletrica.jpg https://ri.ufs.br/jspui/handle/riufs/2442 https://ri.ufs.br/jspui/handle/riufs/26156 Título: Sobre a dependência estatística entre MFCCs e a combinação de energia e frequência fundamental Autor(es): Bezerra, Vitor Magno de Oliveira Santos Abstract: Mel-frequency cepstral coefficients (MFCCs) are a key feature in speech processing. This study investigates the assumption that MFCC values are independent of the combination of energy and fundamental frequency, which carry prosodic information. This investigation is conducted by developing a null hypothesis test to assess the statistical significance of the independence hypothesis between these two sources of information. The results demonstrate that it is statistically implausible for MFCCs to be independent of the combination of energy and fundamental frequency. Furthermore, replicating a speaker recognition system from the literature provides practical evidence of this implausibility, as the results suggest that these two types of features do not complement each other in terms of performance gains. Thu, 20 Aug 2026 00:00:00 GMT https://ri.ufs.br/jspui/handle/riufs/26156 2026-08-20T00:00:00Z https://ri.ufs.br/jspui/handle/riufs/26147 Título: Agregação de variáveis nas redes neurais artificiais baseada em medidas de informação mútua Autor(es): Bastos, Gabriel Francisco Alves Abstract: Artificial neural networks constitute a class of models that lies at the core of mo dern machine learning. A notable attribute of these models is, in many applications, their large number of parameters, which gives rise to the need for either implicit or explicit regularization in order to achieve strong predictive performance. One way to incorporate implicit regularization is by constraining which variables are allowed to interact within artificial neurons, as exemplified by the convolutional and the pooling layers of convolutio nal neural networks (CNNs), where each variable interacts only with its spatial neighbors. Within this context, the central hypothesis motivating this dissertation is that, in certain scenarios, it is beneficial to impose such constraints so that computations performed over subsets of the input variables involve those exhibiting high statistical dependence. In this regard, one of the main contributions of this dissertation is a proof of concept demonstra ting that conventionally defined CNNs provide a practical example of the application of this principle. Furthermore, experimental evidence is presented showing that extending this principle to tabular data through the use of locally connected layers is a promising approach for representation learning and autoencoding. Throughout the dissertation, it is frequently necessary to quantify the statistical dependence between pairs of variables using mutual information measures. Accordingly, this document also presents an in-depth study of the mutual information estimation problem, which remains an active research hotspot. As a result, the outcomes of this study include the proposal of a novel estima tion method for discrete variables and the adoption of a mutual information estimator for continuous sources, which is employed throughout the dissertation. Tue, 18 Aug 2026 00:00:00 GMT https://ri.ufs.br/jspui/handle/riufs/26147 2026-08-18T00:00:00Z https://ri.ufs.br/jspui/handle/riufs/25888 Título: Modelagem por equivalente elétrico para dispositivos a onda acústica de superfície Autor(es): Nogueira, Paulo Gabriel Barreto Abstract: Surface Acoustic Wave (SAW) devices find extensive application in telecommunications as delay lines and highly selective filters, and they also operate as high-sensitivity sensors. This versatility arises from their ability to convert electrical signals into acoustic waves through interdigital transducers and from the strong interaction between these waves and physical or chemical perturbations in the propagation medium. Engineers typically characterize these devices using vector network analyzers, which restrict field applications and complicate integration with systems and wireless sensor networks. These limitations motivate the implementation of approaches based on open- and closed-loop electronic oscillators. However, the development of such systems faces constraints related to the acquisition of SAW devices and to their direct use in circuit simulators. As a result, researchers employ equivalent electrical models, particularly those based on resonator crystals. An extensive literature review reveals that currently available models, although simple and physically implementable, diverge in their representation of the electrical coupling between SAW ports. This aspect plays a critical role in the performance and stability of closed-loop oscillators, especially under uncontrolled environmental conditions. In this context, this work proposes the development of an equivalent electrical circuit model for SAW devices based on a resonator crystal with adjustable parameters. The proposed model more accurately represents the coupling between ports and supports applications in the calibration of SAW-based systems, integration with electronic oscillators, and laboratory testing. This approach reduces dependence on real devices and increases the feasibility of field measurements. The study derives the model from an adapted resonator crystal representation and incorporates dedicated blocks to adjust input and output impedances as well as transmission magnitude. Fri, 06 Mar 2026 00:00:00 GMT https://ri.ufs.br/jspui/handle/riufs/25888 2026-03-06T00:00:00Z https://ri.ufs.br/jspui/handle/riufs/25868 Título: Representação e classificação de perfis térmicos de isoladores poliméricos para diagnóstico de poluição superficial Autor(es): Oliveira, Johnny Herbert Paixão de Abstract: Monitoring insulators in electric power systems is essential for reducing the risk of failures associated with surface pollution. In coastal environments, salt deposition and the presence of moisture can intensify surface leakage current and cause localized heating capable of leading the equipment to failure and causing interruptions in the power supply. In this context, infrared thermography is a promising non-invasive technique because it enables the thermal condition of the equipment to be assessed without direct contact or interruption of operation. This dissertation proposes a methodology for classifying the thermal profiles of polymeric insulators, aiming to distinguish between clean and polluted conditions. To this end, one-dimensional thermal profiles are extracted from thermograms, corrected relative to the ambient temperature, segmented into successive windows, and described using features related to thermal level, temperature dispersion, including variance and standard deviation, contrast, distribution shape, including skewness and kurtosis, spatial variation, and spatial dependence. Three forms of representing the information along the profile are evaluated: aggregated, sequential, and symbolic. The aggregated representation summarizes the features obtained from the windows, whereas the sequential and symbolic representations explicitly preserve the spatial order of these windows. For classification, a support vector machine, a random forest, a long short-term memory recurrent neural network, a hidden Markov model with Gaussian mixtures, and a discrete hidden Markov model are evaluated. The dataset comprises 344 thermal profiles distributed across 69 acquisition groups and includes clean insulators as well as insulators subjected to artificial and natural pollution. Feature selection, model tuning, and the initial performance estimation are conducted through nested validation with separation by acquisition groups, thereby preventing samples originating from the same acquisition from being distributed between the training and test sets. The results indicate that all evaluated models are capable of distinguishing between clean and polluted profiles. Subsequently, a robustness analysis is conducted in which the best models for each representation are evaluated over different repetitions of the training–test split, while keeping fixed the selected features, hyperparameters, and structures defined during nested validation. In this analysis, the support vector machine applied to the aggregated representation achieves the best performance, with a balanced x accuracy of 0.923 ± 0.058 and a multicriteria metric of 0.897 ± 0.067. It is concluded that the spatial segmentation of the thermal profile provides relevant information for identifying surface pollution and constitutes a promising approach for the diagnosis of polymeric insulators. Wed, 05 Aug 2026 00:00:00 GMT https://ri.ufs.br/jspui/handle/riufs/25868 2026-08-05T00:00:00Z