Programa de Pós-Graduação em Engenharia Elétrica e de Computação
Permanent URI for this communityhttps://repositorio.ufrn.br/handle/123456789/11949
Browse
19 results
Search Results
Doctoral Thesis ELSA - Expanded Latent Space Autoencoder Architecture for Feature Extraction: a case study application to Covid-19 time series forecasting(Universidade Federal do Rio Grande do Norte, 2024-09-20) Oliveira, Emerson Vilar de; Gonçalves, Luiz Marcos Garcia; https://orcid.org/0000-0002-7735-5630; http://lattes.cnpq.br/1562357566810393; http://lattes.cnpq.br/8790940901329225; Oliveira, Luiz Affonso Henderson Guedes de; https://orcid.org/0000-0003-2690-1563; http://lattes.cnpq.br/7987212907837941; Silva Júnior, Andouglas Gonçalves da; Santos, Davi Henrique dos; Aroca, Rafael VidalThe global SARS-CoV-2 pandemics compelled governments, institutions, and researchers to assess its impact and develop strategies based on general indicators to achieve the most accurate predictions possible, in order to help managers mitigating its effect. While known epidemiological models were naturally used, they often produced uncertain forecasts due to insufficient or missing data. In addition to data limitation, various machine-learning models such as random forests, support vector regression, LSTM, auto-encoders, and traditional time-series models like Prophet and ARIMA—were employed, yielding impressive yet somewhat limited results. Some of these methods struggle with precision when handling multi-variable inputs, which are crucial for problems like pandemics time series prediction that require both short- and long-term forecasting. In response to this challenge, we propose a novel approach for time-series prediction that utilizes a stacked auto-encoder structure. Our model uses n internal autoencoders to process the input and generate different latent spaces for this respective input. Then these different latent spaces are concatenated and the expanded latent space is obtained. We conducted an experiment using previously published data series on COVID-19 cases, deaths, temperature, humidity, and the air quality index (AQI) in São Paulo City, Brazil. This experiment assessed the suitability of our model for short-, medium-, and long-term forecasting. Furthermore, we directly compared our proposed model with two existing works in the literature that have already undergone expert scrutiny. The first comparison places our model among those that use one network for feature extraction and another for predicting the pandemic trends. The second comparison highlights our model’s effectiveness in multi-series forecasting of pandemic indicators. The results suggest that our proposed model possesses strong capabilities in both feature extraction and multi-series forecasting, offering improvements over the two comparison works. Finally, the model demonstrates promising forecasting accuracy and versatility across datasets of varying lengths, making it a standout option for time-series forecasting tasks.Doctoral Thesis Controle inteligente de sistemas biológicos complexos(Universidade Federal do Rio Grande do Norte, 2024-10-03) Lima, Gabriel da Silva; Bessa, Wallace Moreira; https://orcid.org/0000-0002-0935-7730; http://lattes.cnpq.br/3256782908311485; https://orcid.org/0000-0001-6615-078X; http://lattes.cnpq.br/1393628703238236; Dorea, Carlos Eduardo Trabuco; Araújo, Fábio Meneghetti Ugulino de; Savi, Marcelo Amorim; Cota, Vinicius RosaComplex systems are a class of dynamic systems, typically nonlinear, characterized by the presence of multiple coupled differential equations responsible for describing the dynamic behavior of the system’s internal states. These equations cannot be analyzed in isolation as it would impair the understanding of the system’s overall behavior. Various physiobiological phenomena can be described through complex systems, especially pathologies associated with these phenomena. Controlling this class of dynamic systems allows for the development of techniques that could potentially become new medical treatments in the future. In this work, an intelligent controller is presented, whose main structure is deduced through the Lyapunov Asymptotic Stability Theorem. Embedded into this controller is an adaptive term based on Artificial Neural Networks, implemented to compensate for and predict uncertainties related to unknown model parameters, unmodeled dynamics, and external disturbances. Throughout the text, the controller is tested on different biological complex systems used to represent the brain dynamics of patients with epilepsy and the dynamics of cardiac pathologies. For each example, the controller is also tested in different scenarios where aberrant behaviors of these vital organs may occur. Numerical simulations demonstrate the effectiveness of the controller’s implementation, pointing to a viable path for the development of new treatments beyond the pharmacological area.Master Thesis Controle inteligente de um robô móvel omnidirecional com tomada de decisão utilizando aprendizagem por reforço(Universidade Federal do Rio Grande do Norte, 2021-04-12) Moreira, Victor Ramon Firmo; Bessa, Wallace Moreira; ; ; http://lattes.cnpq.br/5970099237666320; Araújo, Fábio Meneghetti Ugulino de; ; Santana Júnior, Orivaldo Vieira de; ; Colombini, Esther Luna;The evolution of robotic systems has become evident over time. Due to the advances in mechanical manufacturing and the new algorithms used, mobile robots have become increasingly independent in their actions. Regarding machine learning strategies, special attention is given to reinforcement learning algorithms, because of its similarities with the biological learning process. This work proposes the development of an autonomous agent, combining intelligent control strategies with decision-making algorithms. For the implementation of the proposed strategy, the Robotino®omnidirectional mobile robot will be used. Simulations of the robot’s performance were performed to explore space in an environment, for which a specific mathematical model is applied. For system control, the Linearization by Feedback strategy was combined with a compensator based on Artificial Neural Networks to deal with uncertainties, possible external disturbances disturbances and compensate for unmodeled dynamics. The e-greedy algorithm, in turn, was chosen to enable the robot in the decision-making process. The results of the experimental implementation show that an intelligent control strategy was efficient and the proposed intelligent agent was able to explore the environment effectively, obtaining a high average reward.Master Thesis Implementação de um estimador neural de velocidade em um motor de indução trifásico(2019-12-16) Silva, Carlos Yuri Ferreira; Salazar, Andres Ortiz; ; ; Lock, Alberto Soto; ; Araújo, Fábio Meneghetti Ugulino de; ; Paiva, José Álvaro de;This work proposes to perform the speed control of an induction motor (1 cv, 2 poles, 220/380 V) without the use of a mechanical sensor attached to the rotor, avoiding some inconveniences such as a periodic maintenance and increased complexity and number of equipment are always part of the engine running. The three phase induction motor is the most used type of machine for electric drives. Low cost, construction simplicity and reliability are the motives that explains this statement. For the induction motor control was applied the control strategy known as Field Oriented Control and for the estimation of the motor rotor speed was used an artificial neural network, which observer electrical and mechanical system variables. The practical results were satisfatory, since the neural network worked for the measeured and estimated velocity, being possible to perform the control os the motor with the estimated speed in the range of 100 to 170 rad/s.Master Thesis Análise e síntese de superfícies seletivas de frequência bioinspiradas usando redes neurais para aplicações em sistemas de comunicação sem fio(2019-07-01) Paiva, Samuel Belarmino de; D'Assunção, Adaildo Gomes; Silva Neto, Valdemir Praxedes da; ; ; ; D'Assunção Júnior, Adaildo Gomes; ; Fontgalland, Glauco; ; Lins, Hertz Wilton de Castro; ; Costa, José Alfredo Ferreira;This work proposes the analysis of Bioinspired Frequency Selective Surfaces (BFSS) for applications in wireless systems, operating in the C band, Ku band and UWB (ultrawideband). Simple and coupled BFSS structures are considered. The BFSSs have array elements with the four-leaf clover-shaped and maple leaf-shaped, and presented dual-band response, with the operating frequencies in the C and Ku bands. For the development of BFSSs with the ultra-wideband, a cascade structure was developed, in which a FSS with elements with patches with the shape of square loops was coupled to the BFSSs. In addition, the four-leaf clover BFSS synthesis was developed using an artificial neural network with an architecture cascade feedforward and Bayesian regularization training algorithm to obtain the specifications of resonance frequency and respective desired bandwidths. The numerical values obtained by simulations for the projected prototypes were obtained by the ANSYS HFSS software. The prototypes were manufactured and the experimental characterization was performed, whose values were compared to simulation results, which showed good agreement.Master Thesis Análise estatística e técnicas de identificação de pilotos para veículos Baja SAE(2018-08-31) Nunes, Tomaz Filgueira; Martins, Allan de Medeiros; Silva, Ivanovitch Medeiros Dantas da; ; ; ; Dantas, André Felipe Oliveira de Azevedo; ; Nobre, Marcelo Henrique Ramalho;Immersing in the motorsports context, the use of artificial inteligence becomes a great ally to the racing team efficiency by extracting important features from car/driver system and providing feedbacks for a better performance, as it can be found in some Formula 1 teams. From that principle, this work aims to characterize drivers of an off-road Baja SAE vehicle. Through the partnership with the Car-Kará Baja SAE UFRN team, 4 different drivers have been selected in 7 different test tracks. The data has been collected, through an industrial data logger, and analyzed in an offline manner. From the data collection, it has been done a divison of the variable vector in 3 and 5 sections and then it has been computed the statistical analysis for each part, creating the feature vectors. That vector was inserted in an artificial neural archtecture with two hidden layer, obtaining a classification rate of 97% for the variable vector division in 3 parts and 93% for 5 parts.Doctoral Thesis Implementação de uma matriz de neurônios dinamicamente reconfigurável para descrição de topologias de redes neurais artificiais multilayer perceptrons(Universidade Federal do Rio Grande do Norte, 2015-09-04) Silva, Carlos Alberto de Albuquerque; Dória Neto, Adrião Duarte; ; http://lattes.cnpq.br/1987295209521433; ; http://lattes.cnpq.br/7963808444142138; Barbalho, David Simonetti; ; http://lattes.cnpq.br/7208859488227503; Melo, Jorge Dantas de; ; http://lattes.cnpq.br/7325007451912598; Oliveira, José Alberto Nicolau de; ; http://lattes.cnpq.br/2871134011057075; Lopes, Danniel Cavalcante; ; http://lattes.cnpq.br/5342832426660173; Ramos, Karla Darlene Nepomuceno; ; http://lattes.cnpq.br/2751239628595747The Artificial Neural Networks (ANN), which is one of the branches of Artificial Intelligence (AI), are being employed as a solution to many complex problems existing in several areas. To solve these problems, it is essential that its implementation is done in hardware. Among the strategies to be adopted and met during the design phase and implementation of RNAs in hardware, connections between neurons are the ones that need more attention. Recently, are RNAs implemented both in application specific integrated circuits's (Application Specific Integrated Circuits - ASIC) and in integrated circuits configured by the user, like the Field Programmable Gate Array (FPGA), which have the ability to be partially rewritten, at runtime, forming thus a system Partially Reconfigurable (SPR), the use of which provides several advantages, such as flexibility in implementation and cost reduction. It has been noted a considerable increase in the use of FPGAs for implementing ANNs. Given the above, it is proposed to implement an array of reconfigurable neurons for topologies Description of artificial neural network multilayer perceptrons (MLPs) in FPGA, in order to encourage feedback and reuse of neural processors (perceptrons) used in the same area of the circuit. It is further proposed, a communication network capable of performing the reuse of artificial neurons. The architecture of the proposed system will configure various topologies MLPs networks through partial reconfiguration of the FPGA. To allow this flexibility RNAs settings, a set of digital components (datapath), and a controller were developed to execute instructions that define each topology for MLP neural network.Master Thesis Ferramenta para avaliação e inferência de parâmetros de redes industriais sem fio(Universidade Federal do Rio Grande do Norte, 2015-07-20) Florêncio, Heitor Medeiros; Dória Neto, Adrião Duarte; ; http://lattes.cnpq.br/1987295209521433; ; http://lattes.cnpq.br/6422930980833254; Brandão, Dennis; ; http://lattes.cnpq.br/6838931677289559; Silva, Ivanovitch Medeiros Dantas da; ; http://lattes.cnpq.br/3608440944832201; Oliveira, Luiz Affonso Henderson Guedes de; ; http://lattes.cnpq.br/7987212907837941Wireless sensor networks (WSN) have gained ground in the industrial environment, due to the possibility of connecting points of information that were inaccessible to wired networks. However, there are several challenges in the implementation and acceptance of this technology in the industrial environment, one of them the guaranteed availability of information, which can be influenced by various parameters, such as path stability and power consumption of the field device. As such, in this work was developed a tool to evaluate and infer parameters of wireless industrial networks based on the WirelessHART and ISA 100.11a protocols. The tool allows quantitative evaluation, qualitative evaluation and evaluation by inference during a given time of the operating network. The quantitative and qualitative evaluation are based on own definitions of parameters, such as the parameter of stability, or based on descriptive statistics, such as mean, standard deviation and box plots. In the evaluation by inference uses the intelligent technique artificial neural networks to infer some network parameters such as battery life. Finally, it displays the results of use the tool in different scenarios networks, as topologies star and mesh, in order to attest to the importance of tool in evaluation of the behavior of these networks, but also support possible changes or maintenance of the system.Master Thesis Classificador neural híbrido para imagens obtidas por sensoriamento remoto(Universidade Federal do Rio Grande do Norte, 2011-08-12) Lima, Alexandre Gomes de; Guerreiro, Ana Maria Guimarães; ; http://lattes.cnpq.br/8556144121380013; ; http://lattes.cnpq.br/4063478137671603; Dória Neto, Adrião Duarte; ; http://lattes.cnpq.br/1987295209521433; Henriques, Antônio de Pádua de Miranda; ; http://lattes.cnpq.br/9855577471019220; Soares, Heliana Bezerra;Remote sensing is one technology of extreme importance, allowing capture of data from the Earth's surface that are used with various purposes, including, environmental monitoring, tracking usage of natural resources, geological prospecting and monitoring of disasters. One of the main applications of remote sensing is the generation of thematic maps and subsequent survey of areas from images generated by orbital or sub-orbital sensors. Pattern classification methods are used in the implementation of computational routines to automate this activity. Artificial neural networks present themselves as viable alternatives to traditional statistical classifiers, mainly for applications whose data show high dimensionality as those from hyperspectral sensors. This work main goal is to develop a classiffier based on neural networks radial basis function and Growing Neural Gas, which presents some advantages over using individual neural networks. The main idea is to use Growing Neural Gas's incremental characteristics to determine the radial basis function network's quantity and choice of centers in order to obtain a highly effective classiffier. To demonstrate the performance of the classiffier three studies case are presented along with the results.Master Thesis Sistema não-Intrusivo para Estimação da Direção do olhar utilizando redes neurais artificiais(Universidade Federal do Rio Grande do Norte, 2010-11-26) Peixoto, Helton Maia; Guerreiro, Ana Maria Guimarães; Dória Neto, Adrião Duarte; ; http://lattes.cnpq.br/1987295209521433; ; http://lattes.cnpq.br/8556144121380013; ; http://lattes.cnpq.br/8709900833456787; Soares, Heliana Bezerra; ; Santos, Celso Alberto Saibel; ; http://lattes.cnpq.br/7614206164174151The fundamental senses of the human body are: vision, hearing, touch, taste and smell. These senses are the functions that provide our relationship with the environment. The vision serves as a sensory receptor responsible for obtaining information from the outside world that will be sent to the brain. The gaze reflects its attention, intention and interest. Therefore, the estimation of gaze direction, using computer tools, provides a promising alternative to improve the capacity of human-computer interaction, mainly with respect to those people who suffer from motor deficiencies. Thus, the objective of this work is to present a non-intrusive system that basically uses a personal computer and a low cost webcam, combined with the use of digital image processing techniques, Wavelets transforms and pattern recognition, such as artificial neural network models, resulting in a complete system that performs since the image acquisition (including face detection and eye tracking) to the estimation of gaze direction. The obtained results show the feasibility of the proposed system, as well as several feature advantages.