Programa de Pós-graduação em Engenharia Elétrica e da Computação
URI Permanente desta comunidade
Navegar
Navegando Programa de Pós-graduação em Engenharia Elétrica e da Computação por Por Orientador "CRUZ JÚNIOR, Gélson da"
Agora exibindo 1 - 4 de 4
Resultados por página
Opções de Ordenação
Item Previsão de Vazões Naturais Diárias Afluentes ao Reservatório da UHE Tucuruí Utilizando a Técnica de Redes Neurais Artificiais(Universidade Federal de Goiás, 2012-09-05) FERREIRA, Carlos da Costa; CRUZ JÚNIOR, Gélson da; http://lattes.cnpq.br/4370555454162131The forecast of natural flows to hydroelectric plant reservoirs is an essential input to the planning and programming of the SIN´s operation. Various computer models are used to determine these forecasts, including physical models, statistical models and the ones developed with the RNA´s techniques. Currently, the ONS performs daily forecasts of natural flows to the UHE Tucuruí based on the univariate stochastic model named PREVIVAZH, developed by Electric Energy Research Center - Eletrobras CEPEL. Throughout the last decade, several papers have shown evolution in the application of neural networks methodology in many areas, specially in the prediction of flows on a daily, weekly and monthly basis. The goal of this dissertation is to present and calibrate a model of natural flow forecast using the RNA´s methodology, more specifically the NSRBN (Non-Linear Sigmoidal Regression Blocks Networks) (VALENCA; LUDERMIR, 2001), on a time lapse from 1 to 12 days forward to the Tucuruí Hydroelectric Plant, considering the hydrometric stations data located upstream from it s reservoir. In addition, a comparative analysis of results found throughout the calibrated neural network and the ones released by ONS is performed. The results show the advantage of the methodology of artificial neural networks on autoregressive models. The Mean Absolute Percentage Error - MAPE values obtained were, on average, 48 % lower than those released by the ONS.Item Estudo de Técnicas de Otimização de Sistemas Hidrotérmicos por Enxame de Partículas(Universidade Federal de Goiás, 2012-06-21) GOMIDES, Lauro Ramon; CRUZ JÚNIOR, Gélson da; http://lattes.cnpq.br/4370555454162131Particle Swarm Optimization has been widely used to solve real-world problems, including the operation planning of hydrothermal generation systems, where the main goal is to achieve rational strategies of operation. This can be accomplished by minimizing the high-cost thermoelectric generation, while maximizing the low-cost hydroelectric generation. The optimization process must consider a set of complex constrains. This work presents the application of some recently proposed Particle Swarm Optimizers for a group of hydroelectric power plants of the Brazilian interconnected system, using real data from existing plants. There were performed some tests by using the standard PSO, PSO-TVAC, Clan PSO, Clan PSO with migration, Center PSO, and one approach proposed in this work, called Center Clan PSO, over three different mid-term periods. All PSO approaches were compared to the results achieved by a Non-linear Programming algorithm (NLP). Furthermore, another approach was proposed, based on Center PSO, named Extended Center PSO. It was observed that the PSO approaches presented as promising solutions to the problem, even better than NLP in some cases.Item Análise do modelo de Kiyotaki-Wright em simulações multiagentes que utilizam a plataforma Swarm(Universidade Federal de Goiás, 2011-02-10) RODOVALHO, Wildener Monteiro; VINHAL, Cássio Dener Noronha; http://lattes.cnpq.br/9791117638583664; CRUZ JÚNIOR, Gélson da; http://lattes.cnpq.br/4370555454162131The goal of this dissertation is to make a methodological repetition of Duffy and Ochs's and Rouchier's works. Nonetheless, it tries to have a different focus. Among many theoreticaleconomical models which explain the emergence of money, there is one that distinguishes itself due to its simplicity, efficiency, and robustness: Kiyotaki-Wright's model. Therefore, some economists devoted themselves to apply this model in experiments with human beings and in computer simulations. Among those economists, Duffy and Ochs applied it later, correcting some flaws of the application process used by their predecessors. Duffy and Ochs's studies showed that Kiyotaki-Wright's model is actually valid for its purpose. However, besides all the efforts made, their agents were not able to achieve the foreseen equilibrium. In the same way, the French researcher Rouchier repeated methodologically the works of Duffy and Ochs while criticizing their implementation details. From this research, Rouchier creates other four sub-models in order to take her agents to the proposed equilibrium, obtaining partial success. In this dissertation, the computer system development platform is Swarm, especially created to multi-agent simulations and used, most of the time, in social and biological scope models. The agents' rationalization process is undertaken in trial and error way in accordance with the reinforcement learning technique, analyzing the answer given by the environment in each period. Finally, a new entity called Mediator is created to intermediate the relations between agents and mainly to pair them randomly. The obtained results are, in general, closer to the equilibrium when compared to the results of previous research. From them, some studies are made about the relation between the number of agents and the number of periods of simulation as values approximates the equilibrium.Item Análise e Comparação de Modelos de Previsão de Vazões para o Planejamento Energético, Utilizando Séries Temporais(Universidade Federal de Goiás, 2009-01-02) XAVIER, Priscila Branquinho; CRUZ JÚNIOR, Gélson da; http://lattes.cnpq.br/4370555454162131n the planning of the energetic operation, analysis and forecasts of the flow are very important. A huge difficulty in the forecast of flow is the seasonality presence, due to drought and flood periods in the year. Many scientists, with different methodologies, have been concerned with finding a best model, compared with the utilized by Brazil s system - Markovian Model. The Makovian Model, or selfregressive with order 1, is a Box & Jenkins methodology, and requires data handling to treat non-stationarity, or the use of regular models, requiring a hardly theoretical formulation for the statistical procedures. Therefore, the statistical models, autoregressive model with seasonality and Holt-Winters model, of treatment of temporal series are presented and, carried out the flow s analysis and forecast for three study groups, in two different (historical) horizons. The performance of the models was compared and the results showed that the proposed models presents better adjust than the model adopted by Brazilian system