https://ri.ufs.br/jspui/handle/riufs/2495 2026-04-26T21:52:45Z https://ri.ufs.br/jspui/handle/riufs/21564 Título: Sobre a capacidade de representação de um nariz eletrônico composto por dois sensores do tipo óxido metálico Autor(es): Santos, Ítalo de Oliveira Abstract: The use of electronic noses for the analysis of odor signals has been increasing over the last decades. From applications in food discrimination and quality control to medical applications, many studies have used these devices and studied the characteristics associated with the sensors used in their construction. Metal oxide (MOX) sensors are one of the most widely used, but they still have disadvantages that continue to be studied and compensated for in some way, such as their low selectivity. In this context, this work aimed to propose a quantifiable way, based on the knowledge of information theory, to compare the capacity of spaces generated by simultaneous responses of MOX sensors to represent, with an arbitrary low classification error, different combinations of gas mixtures present in an environment with an arbitrarily low error. The estimation of this proposed quantity is demonstrated for simulated data from an array of two different MOX sensors based on a commercial sensor operating at two different temperature modulation levels, in which a representation capacity of a few dozen classes for the signal space was observed, when measurements were analyzed in a steady state without drift effect, under moderate noise levels. It is believed that this proposed quantity may be useful to further understand and compare these devices that constitute electronic noses to establish theoretical limits regarding their use in classifiers. 2025-03-12T00:00:00Z https://ri.ufs.br/jspui/handle/riufs/21557 Título: Oscilador H para medição de capacitâncias Autor(es): Carvalho, Stéphane Santos Abstract: Capacitive sensors operate based on capacitance variation in response to a stimulus, offering a wide range of applications. Among the various measurement techniques, oscillator circuits stand out for providing a stable, direct, and reliable response. However, in applications where isolating the sensor electrodes is not feasible, the presence of a resistive component associated with capacitance—often overlooked—can compromise oscillator-based measurement systems by destabilizing the circuit and interrupting its operation, especially at high frequencies. Although the literature addresses this issue, most solutions focus solely on low-frequency applications and propose large, complex circuits, hindering the development of efficient systems and limiting the use of high-frequency sensors. In light of this, this work presents a new sinusoidal oscillator topology, based on the Colpitts oscillator, for measuring non-ideal capacitive loads in the frequency range from hundreds of kilohertz to hundreds of megahertz, referred to here as the H Oscillator. This topology is designed to maintain oscillation criteria even in the presence of variations in the sensitive elements. The scientific framework adopted for the oscillator’s development, as well as its theoretical formulation, are detailed in this work. Simulations were performed using LTSpice and QucsStudio simulation environments, employing capacitive sensor models. The results demonstrate the circuit’s robustness against parameter variations, enabling the measurement of different capacitive loads without oscillation interruption. Furthermore, the theoretical formulation was validated, with errors of only 0.78% between the expected and measured frequency variation curves and 1.47% for amplitude variation. Thus, this dissertation presents a novel approach to capacitive load sensor measurement, with potential for various applications. 2025-02-18T00:00:00Z https://ri.ufs.br/jspui/handle/riufs/21552 Título: Um estudo sobre a extração de características em dados do Twitter na tarefa de detecção de depressão Autor(es): Santos, Ataíde Mateus Gualberto dos Abstract: Depression is a mental condition that affects millions of people worldwide, manifesting through persistent feelings of sadness, lack of interest, and changes in thought and behavior patterns. With the rise of social media, it has become possible to identify signs of depression through users' posts, offering new opportunities for the study of linguistic indicators associated with this condition. This study explores methods for feature extraction in social media data, aiming to identify signs of depression in Twitter users. The research began with the creation of a database from public posts, followed by the application of data preprocessing techniques. Cognitive Behavioral Theory was integrated into the theoretical framework, providing the basis for manual feature extraction. The selection of the most relevant features was carried out through hypothesis testing combined with the AdaBoost classifier. Among the key indicators found, the frequent use of first-person words and an increase in posts during nighttime by individuals labeled as depressed, compared to the control group, stood out. Additionally, data analysis revealed a lower number of effective words in the vocabulary of those labeled with depression. The results of this study have the potential to contribute to society. Health professionals can use data-driven screening tools and guide the creation of public mental health policies. Moreover, these techniques can assist social media platforms in identifying and supporting users at risk. 2024-08-26T00:00:00Z https://ri.ufs.br/jspui/handle/riufs/20509 Título: Monitoramento automatizado de isolamentos por imageamento ultravioleta empregando aprendizado profundo Autor(es): Rodrigues, Gustavo Aragão Abstract: The presence of surface discharges or corona discharges in the vicinity of equipment and components of the electrical system is, in general, an indication of the occurrence of some undesirable phenomenon. In many cases, it potentially indicates a process that can lead to the failure or physical degradation of materials. One of the most promising techniques of corona effect monitoring is the use of specialized cameras for the detection of ultraviolet radiation. This dissertation presents an innovative algorithm for classifying the criticality of insulation based on attributes extracted from videos recorded with a camera capable of detecting ultraviolet radiation. The proposed methodology is based on extracting three attributes from each detected facula origin: maximum persistence, area and minimum distance between facula origin and isolation. To obtain this last attribute, we proposed a methodology for segmenting insulation in images using a combination of a deep convolutional neural network model and an adaptive thresholding method based on the mean. The deep learning model achieved 85.5% precision in detecting insulation on a validation dataset consisting of 1985 images and 8730 instances. The classification results showed that the distance between insulation and facula origin is an essential attribute for video analysis. This variable provides context for recorded discharges and allows differentiation between cases where ultraviolet radiation originates from insulation and those where discharge location is less critical. 2023-08-22T00:00:00Z