Programa de Pós-Graduação em Ciência da Computação - PPGCC/ICEN

URI Permanente desta comunidadehttps://repositorio.ufpa.br/handle/2011/2351

O Programa de Pós-Graduação em Ciência da Computação (PPGCC) é um programa do Instituto de Ciências Exatas e Naturais (ICEN) da Universidade Federal do Pará (UFPA). O PPGCC foi fundando em 2005 e visa ampliar a massa crítica na região, objetivando a produção de pesquisa científica na área, na busca de resolução de problemas regionais.

Navegar

Agora exibindo 1 - 9 de 9

Acesso aberto (Open Access)
Aplicação de sensores virtuais na estimação da concentração dos parâmetros físico-químicos e metais em corpos d’água de reservatórios de hidrelétricas: um estudo de caso na Região Amazônica
(Universidade Federal do Pará, 2014-10-23) RIBEIRO NETO, Benedito de Souza; OLIVEIRA, Terezinha Ferreira de; http://lattes.cnpq.br/6230804143692945; SILVEIRA, Antônio Morais da; http://lattes.cnpq.br/7549503749842625
This research introduces the use of virtual sensors to estimate the concentration of physico-chemical parameters and metals in monitoring water quality of reservoirs Amazon through artificial neural networks (ANN) and images of remote sensing. A factor analysis of the variables considered in the study confirmed the relationship of the first factor with Secchi disk, Total Iron, PO4, Total P, TSS and Turbidity on a single factor, as these have a high reflectance and good energy absorption by satellite sensors. These elements were determined by ANN's, producing satisfactory results approach 100% between observed and estimated. The tests resulted in a good approximation, the first band Secchi disk depth, total Fe, STS, and turbidity of the water reservoir. In the specific case of the parameters PO4 and Total P, besides the problem of the small number of sampling stations available data and the variability inherent in the hydrological cycle of the region, it was found, through the interpretation of images, lack of similarities between the data used in training and validation of RNA. Overall, the study demonstrated the effectiveness of the application of virtual sensors in monitoring water quality of reservoirs in the Amazon by satellite imagery, providing a precise and less expensive alternative resources in the process of environmental monitoring.
Acesso aberto (Open Access)
Aplicação de técnicas estatísticas e de inteligência computacional na classificação de ciclos hidrológicos em reservatórios de água na região amazônica: um estudo de caso
(Universidade Federal do Pará, 2014-05-09) FREIRE, Jean Carlos Arouche; OLIVEIRA, Terezinha Ferreira de; http://lattes.cnpq.br/6230804143692945; MORAIS, Jefferson Magalhães de; http://lattes.cnpq.br/5219735119295290
This study evaluates the quality of the water reservoir of the Hydroelectric Plant Tucuruí according to the regional hydrological cycle and the spatial arrangement of the different sampling sites distributed in areas upstream of the dam in the period 2009-2012 from the amendment of 17 parameters physico-chemical and metals from water extracted of six factors that accounted for 71.01% of total data variability. It was observed that the greatest variations of NO3, NH4, , totalP, PO4 and STS occurred in the period of floods and may be an indication of trophic status in the sampling sites due to the existence of fishing poles or population density in the vicinity these sites. Discriminant analysis, artificial neural networks, k-nearest neighbors, support vector machine with polynomial and radial core and random forest: classification of the hydrological cycle to six classifiers were used. The results indicate that the random forest classifier showed the best performance with a percentage rating of 7.80% of incorrect predictions. While Student t test indicates that random forest and k-nearest neighbors have an average rate of incorrect predictions with equal significance index set at α = 5%.
Acesso aberto (Open Access)
Avaliação de desempenho de técnicas de localização em ambientes reais aplicadas a redes de sensores sem fio
(Universidade Federal do Pará, 2014-05-26) MACHADO, Leomário Silva; MONTEIRO, Dionne Cavalcante; http://lattes.cnpq.br/4423219093583221; ARAÚJO, Josivaldo de Souza; http://lattes.cnpq.br/8158963767870649
The location of wireless sensor networks is a challenge that goes beyond the use of popular GPS through several studies that aim to improve it or even replace it. The location can be performed using multiple antennas and their respective angles, and time synchronization, time differential between transmission of two different or same power with a radio signal. From these patterns estimated, various techniques have been postulated in order to use the resources available to measure distances and estimate the coordinates of a node. Among these techniques may be cited as the most important Lateração, Nearest Neighbor, K-Nearest Neighbor, Min-Max, Non-Linear Regression, Linear Regression Non-Iterative, Sum-Dist, Dv-hop, Artificial Neural Network, filter Kalman . This paper conducts a series of tests conducted in two environments, the first indoor, outdoor and the second using the hardware as the MEMSIC IRIS modules to perform the experiment. These tests are compared Lateração, KNN and an Artificial Neural Network techniques is proposed for the purpose of estimating the location of a WSN node. Lateração mathematical formulations KNN and are presented as well as the configuration of the neural network used in the tests conducted in this work. The results are shown taking the benchmark for comparative analysis techniques to the percentage there between and better quantitative analysis, the data are tabulated for display accuracy.
Acesso aberto (Open Access)
Controle energeticamente eficiente de múltiplos saltos para redes de sensores sem fio heterogêneas utilizando lógica fuzzy
(Universidade Federal do Pará, 2014-03-21) SILVA, Alexandre Márcio Melo da; SILVEIRA, Antônio Morais da; http://lattes.cnpq.br/7549503749842625
This study presents a centralized control to elect appropriate Cluster Heads (CHs), assuming three levels of heterogeneity and multi-hop communication between Cluster Heads. The centralized control uses the k-means algorithm, responsible for the division of clusters and Fuzzy Logic to elect the Cluster Head and selecting the best route of communication between elected. The results indicate that the proposal offers great advantages, allowing us to select the most suitable nodes for group leaders at each round based on the Fuzzy System values, and also the use of Fuzzy Logic as a decision tool to implement multiple hops between CHs, since it minimizes the power dissipation of the selected CHs more distant from the collection point. The insertion of three levels of heterogeneity,corresponding to normal, advanced and super sensors, contributes considerably to increasing the period of network stability. Another great contribution obtained from the is the use of a central control in base station (BS) with advantages over local information processing in each node, a process usually found in traditional algorithms for electing CHs. The proposed solution proved that the election of the more efficient CH, considering its location and energy levels discrepancies, and also, the inclusion of new heterogeneity levels, allows to increase the networkstability period, ie, the period that the network is fully functional, greatly increasing the useful lifetime in heterogeneous WSN.
Acesso aberto (Open Access)
Um método fuzzy para programação de inspeções em transformadores baseado na análise de gases combustíveis e parâmetros físico-químicos do óleo isolante
(Universidade Federal do Pará, 2014-10-24) PANTOJA, Márcia Priscila Furtado; OLIVEIRA, Terezinha Ferreira de; http://lattes.cnpq.br/6230804143692945; SILVEIRA, Antônio Morais da; http://lattes.cnpq.br/7549503749842625
Power transformers using oil for insulation internal parts need periodic maintenance of a program aimed at detecting possible failures, such as deterioration of the winding insulation cellulose, short circuit between turns, gases dissolved in the oil due to its degeneration etc. Typically, preventive maintenance programs are used by companies for generation, transmission and distribution of electricity for avoiding these failures may result in removal of equipment from the electrical system, a situation that involves large financial losses, either by decreased billing at payment of penalties to the regulatory agency, as well as reduced system reliability. To avoid these types of losses, this paper presents a proposed solution using fuzzy approach, to assist the process of inspections programming in high power transformers (500kv), through the analysis of gases dissolved in insulating oil, obtained by the technique gas chromatography and physico-chemical analysis of the oil. We attempted to produce a solution that combines the results of traditional methods already established in the literature, with additional situations originating from the physico-chemical analysis of the oil and knowledge of experts in order to increase the effectiveness of such equipment inspection procedure. With this, any Electric Power System - SEP - that performs diagnostics on transformers by NBR 7274, tends to get a more accurate diagnosis, due to the association of physical-chemical analysis of the insulating oil with gas chromatography and knowledge tacit from specialists, enabling us to expand the base of original rules of that technical standard. All tests and the database used were from a laboratory of Chemistry of various electrical substations of an power company located in Barcarena - Pará - Brazil.
Acesso aberto (Open Access)
Mineração de dados educacionais: relato de experiência no ambiente virtual LABSQL
(Universidade Federal do Pará, 2014-09-19) DIAS, Maxwel Macedo; FAVERO, Eloi Luiz; http://lattes.cnpq.br/1497269209026542
One of the digital technologies used in current practices of online education is the Virtual Learning Environment (VLE). Educators use these environments to provide online information, but have little support to assess the learning of learners at a distance, so that the lack of perception of the educator as to the state of understanding of their students can lead tofailure of an online course. Most VLEs store large volumes of data from the history of accesses to system resources made by students, their assessments, among others. In recent years the Educational Data Mining has been used to explore the data from educational settings, as well as better understand the students and their teaching and learning. The objective of this work is to assess online learning by the use of Educational Data Mining on the LabSQL virtual environment used in the Federal University of Pará, through the application of techniques called Decision Tree, Bayesian Network, Association Rules and Cluster Analysis. The results obtained proved to be efficient to support educators in the assessment of online learning because they allow the analysis of student’s profile regarding the use of this technology and the teaching-learning environment in LabSQL. Moreover, the rules generated from data mining indicate how the student can improve learning by better using the environment.
Acesso aberto (Open Access)
Uma ontologia que estabelece os relacionamentos de dependência entre as práticas de gerência de requisitos e gerência de projetos constantes nos modelos MR-MPS-SW e CMMI-DEV
(Universidade Federal do Pará, 2014-04-16) YOSHIDOME, Ewelton Yoshio Chiba; OLIVEIRA, Sandro Ronaldo Bezerra; http://lattes.cnpq.br/2080791630485427
The models and standards for software quality aim to propose a set of best practices to support the software companies in order to achieve the continuous software process improvements, so, those companies may develop better products. Commonly, the software companies need to hire implementation consultants in order to support the best practices institutionalization. In this context, there are companies and novice consultants that have the difficulty to understand about the practices and the relationship between those practices in quality models. In order to resolve it, we will create a ontology. The ontology is “a set of ordered terms, by hierarchy, in order to describe a domain that can be used as a knowledge base”. Thus, this research proposes a ontology that establish the dependency relationship among the best practices of project management process and requirement management process. In order to achieve this result, we need: (1) a study about the best practices presents in MR-MPS-SW and CMMI-DEV models, in order to find the relationship among the best practices; (2) create a ontology in UML – Unified Modeling Language; (3) define the ontology axioms, in order to consolidate the meaning of dependency relationships; and (4) survey the companies that have official certification, in this survey we collected the evidences used to implement the best practices suggested in quality models, in order to instantiate the ontology for its evaluation. In the end, our outcome was a ontology (structured model and axioms) that establish the relationship of the best practices among Project Management and Requirement Management process.
Acesso aberto (Open Access)
Otimização por multi-enxame evolucionário de partículas clássico e quântico competitivo sob a arquitetura paralela CUDA aplicado em problemas de engenharia
(Universidade Federal do Pará, 2014-05-23) SOUZA, Daniel Leal; OLIVEIRA, Roberto Célio Limão de; http://lattes.cnpq.br/4497607460894318; MONTEIRO, Dionne Cavalcante; http://lattes.cnpq.br/4423219093583221
This paper presents the development of a set of hybrid metaheuristic based on the use of evolutionary strategies in conjunction with classical and quantum multi-swarm optimization with master-slave approach. These algorithms are named Competitive Evolutionary Multi-Swarm Optimization (CEMSO) and Competitive Quantum-Behaviour Evolutionary Multi-Swarm Optimization (CQEMSO). For comparison and validation of the results, four engineering problems encountered in many publications scientific are used: Welded Beam Design (WBD); Minimization of the Weight of a Tension/ Compression Spring (MWTCS); Speed Reducer Design (SRD); Design of a Pressure Vessel (DPV). The algorithms were developed under the CUDA architecture, which provides a massive parallel computing environment that enables a more appropriate data allocation regarding the organization of swarms, as well as contributing to the significant decrease in processing time. With the application of evolutionary strategies in the PSO and QPSO algorithms, as well as the proposed boundary conditions, the solutions described in this document offer several advantages. We can highlight improvements in the ability to search, increasing the convergence rate and high degree of parallelism. These facts are confirmed by the data obtained (i.e. Execution time, best solutions obtained, mean and variance of results) by CQEMSO and CQEMSO algorithms when compared to those obtained from multi-swarm approach for PSO (COMSO), EPSO (COEMSO) and QPSO (COQMSO). All of these algorithms were implemented and subjected to performance analysis through experiments with engineering problems described above.
Acesso aberto (Open Access)
Reposdn: um método de organização e coordenação de aplicações em repositório para redes definidas por software
(Universidade Federal do Pará, 2014-02-28) PINHEIRO, Rômulo Silva; ABELÉM, Antônio Jorge Gomes; http://lattes.cnpq.br/5376253015721742
The adoption of the paradigm of Software Defined Networks - SDN by academic institutions and large companies like Google, Cisco, HP, among others, is already a reality. This paradigm is considered a solution to the new model of Future Internet due to its technical and economic viability. The SDN paradigm enables new perspectives in terms of abstractions, control and development applications, free from the limitations of current technologies environments. However there is still lack in the way the management and control of several existing applications in the Internet occurs. The programmability is key to successful paradigm for Software Defined Networks as it allows the separation of the data with the control plane level. However, in the context of this paradigm, manage input and output controllers for many applications is a relatively unexplored issue, since the SDN applications, routing applications such as, for example, are not organized in a controlled environment. This thesis proposes a method of organization and coordination applications in repository for software defined networks, called RepoSDN, which specifies, models and automates the processes needed for applications reaching the administrator can manage this and their uses in the network, making it simple, safe and agile. The RepoSDN presents a structure where each SDN application is managed within the network to reach out to controllers. Additionally, it contains a central repository, where any user can send their applications to share with management Software Defined Network, which use specific applications for your need. A prototype of RepoSDN, which implements the functionality of the applications repository and components designed to help in the management of applications on a programmable network was developed. The method of organization and coordination of applications repository for SDN helps the administrator to handle each application. The proposal has been validated through the development of the repository of application, which consisted in doing load testing in developed component testing. The objective of the components is becoming simpler method that applications are entered in the various types of controllers of Software Defined Networking, resulting in the proof of feasibility, scalability and exibility of RepoSDN.

Navegar

Navegando Programa de Pós-Graduação em Ciência da Computação - PPGCC/ICEN por CNPq "CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO"