840 resultados para Equalização Adaptativa. Redes Neurais. Sistemas Ópticos. Equalizador Neural


Relevância:

100.00% 100.00%

Publicador:

Resumo:

The aim of this work is to advance a new approach for estimating demographic density, through combining a Geographic Information System with GMDH Neural Networks. The model that is suggested parts the analyzed space into a rectangular grid formed by multiple cells measuring 0.01 km2 each. The forecasts are elaborated based on the demographic density in each cell and in its neighboring cells at a given time. Despite the limited availability of data during the modeling phase, the utilization of this method for studying a Brazilian medium-sized city presented promising results.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

As sintonias dos Controladores PID existentes em um Sistema de Posicionamento Dinâmico, utilizado em embarcações e plataformas a fim de manter uma posição fixa em alto-mar ou de realizar determinada manobra, sempre tem sido um desafio a ser vencido. Trata-se de uma tarefa demorada, dependente das condições ambientais e com um elevado custo financeiro, uma vez que as horas dedicadas do profissional habilitado são caras. Além disso, a embarcação deve-se manter estabilizada durante o período de tempo no qual determinada função é realizada, como por exemplo, perfuração, abastecimento, ou lançamento de dutos. Foi utilizado um software para simular o posicionamento de uma embarcação em alto-mar sob diversas condições de vento e correnteza, com o qual foi possível verificar a influência da sintonia dos parâmetros PID do Controlador no desempenho do sistema de controle. O Sistema dinâmico abordado possui um comportamento não linear e sujeito a fortes distúrbios não medidos, o que são apenas alguns exemplos de questões avaliadas deste trabalho. Neste contexto, foram projetadas Redes Neurais com o intuito de aprimorar a técnica utilizada para determinar os ganhos de um dos Controladores PID de um Sistema de Posicionamento Dinâmico. Os melhores resultados foram obtidos através da avaliação de desempenho de diversas simulações de Redes Neurais que revelam a viabilidade da implementação da sintonia automática de Controladores em Sistemas de Posicionamento Dinâmico.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

A capacidade de encontrar e aprender as melhores trajetórias que levam a um determinado objetivo proposto num ambiente e uma característica comum a maioria dos organismos que se movimentam. Dentre outras, essa e uma das capacidades que têm sido bastante estudadas nas ultimas décadas. Uma consequência direta deste estudo e a sua aplicação em sistemas artificiais capazes de se movimentar de maneira inteligente nos mais variados tipos de ambientes. Neste trabalho, realizamos uma abordagem múltipla do problema, onde procuramos estabelecer nexos entre modelos fisiológicos, baseados no conhecimento biológico disponível, e modelos de âmbito mais prático, como aqueles existentes na área da ciência da computação, mais especificamente da robótica. Os modelos estudados foram o aprendizado biológico baseado em células de posição e o método das funções potencias para planejamento de trajetórias. O objetivo nosso era unificar as duas idéias num formalismo de redes neurais. O processo de aprendizado de trajetórias pode ser simplificado e equacionado em um modelo matemático que pode ser utilizado no projeto de sistemas de navegação autônomos. Analisando o modelo de Blum e Abbott para navegação com células de posição, mostramos que o problema pode ser formulado como uma problema de aprendizado não-supervisionado onde a estatística de movimentação no meio passa ser o ingrediente principal. Demonstramos também que a probabilidade de ocupação de um determinado ponto no ambiente pode ser visto como um potencial que tem a propriedade de não apresentar mínimos locais, o que o torna equivalente ao potencial usado em técnicas de robótica como a das funções potencias. Formas de otimização do aprendizado no contexto deste modelo foram investigadas. No âmbito do armazenamento de múltiplos mapas de navegação, mostramos que e possível projetar uma rede neural capaz de armazenar e recuperar mapas navegacionais para diferentes ambientes usando o fato que um mapa de navegação pode ser descrito como o gradiente de uma função harmônica. A grande vantagem desta abordagem e que, apesar do baixo número de sinapses, o desempenho da rede e muito bom. Finalmente, estudamos a forma de um potencial que minimiza o tempo necessário para alcançar um objetivo proposto no ambiente. Para isso propomos o problema de navegação de um robô como sendo uma partícula difundindo em uma superfície potencial com um único ponto de mínimo. O nível de erro deste sistema pode ser modelado como uma temperatura. Os resultados mostram que superfície potencial tem uma estrutura ramificada.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Este trabalho tem por motivação evidenciar a eficiência de redes neurais na classificação de rentabilidade futura de empresas, e desta forma, prover suporte para o desenvolvimento de sistemas de apoio a tomada de decisão de investimentos. Para serem comparados com o modelo de redes neurais, foram escolhidos o modelo clássico de regressão linear múltipla, como referência mínima, e o de regressão logística ordenada, como marca comparativa de desempenho (benchmark). Neste texto, extraímos dados financeiros e contábeis das 1000 melhores empresas listadas, anualmente, entre 1996 e 2006, na publicação Melhores e Maiores – Exame (Editora Abril). Os três modelos foram construídos tendo como base as informações das empresas entre 1996 e 2005. Dadas as informações de 2005 para estimar a classificação das empresas em 2006, os resultados dos três modelos foram comparados com as classificações observadas em 2006, e o modelo de redes neurais gerou o melhor resultado.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

O presente trabalho apresenta uma nova metodologia de localização de faltas em sistemas de distribuição de energia. O esquema proposto é capaz de obter uma estimativa precisa da localização tanto de faltas sólidas e lineares quanto de faltas de alta impedância. Esta última classe de faltas representa um grande problema para as concessionárias distribuidoras de energia elétrica, uma vez que seus efeitos nem sempre são detectados pelos dispositivos de proteção utilizados. Os algoritmos de localização de faltas normalmente presentes em relés de proteção digitais são formulados para faltas sólidas ou com baixa resistência de falta. Sendo assim, sua aplicação para localização de faltas de alta impedância resulta em estimativas errôneas da distância de falta. A metodologia proposta visa superar esta deficiência dos algoritmos de localização tradicionais através da criação de um algoritmo baseado em redes neurais artificiais que poderá ser adicionado como uma rotina adicional de um relé de proteção digital. O esquema proposto utiliza dados oscilográficos pré e pós-falta que são processados de modo que sua localização possa ser estimada através de um conjunto de características extraídas dos sinais de tensão e corrente. Este conjunto de características é classificado pelas redes neurais artificiais de cuja saída resulta um valor relativo a distância de falta. Além da metodologia proposta, duas metodologias para localização de faltas foram implementadas, possibilitando a obtenção de resultados comparativos. Os dados de falta necessários foram obtidos através de centenas de simulações computacionais de um modelo de alimentador radial de distribuição. Os resultados obtidos demonstram a viabilidade do uso da metodologia proposta para localização de faltas em sistemas de distribuição de energia, especialmente faltas de alta impedância.

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Forecast is the basis for making strategic, tactical and operational business decisions. In financial economics, several techniques have been used to predict the behavior of assets over the past decades.Thus, there are several methods to assist in the task of time series forecasting, however, conventional modeling techniques such as statistical models and those based on theoretical mathematical models have produced unsatisfactory predictions, increasing the number of studies in more advanced methods of prediction. Among these, the Artificial Neural Networks (ANN) are a relatively new and promising method for predicting business that shows a technique that has caused much interest in the financial environment and has been used successfully in a wide variety of financial modeling systems applications, in many cases proving its superiority over the statistical models ARIMA-GARCH. In this context, this study aimed to examine whether the ANNs are a more appropriate method for predicting the behavior of Indices in Capital Markets than the traditional methods of time series analysis. For this purpose we developed an quantitative study, from financial economic indices, and developed two models of RNA-type feedfoward supervised learning, whose structures consisted of 20 data in the input layer, 90 neurons in one hidden layer and one given as the output layer (Ibovespa). These models used backpropagation, an input activation function based on the tangent sigmoid and a linear output function. Since the aim of analyzing the adherence of the Method of Artificial Neural Networks to carry out predictions of the Ibovespa, we chose to perform this analysis by comparing results between this and Time Series Predictive Model GARCH, developing a GARCH model (1.1).Once applied both methods (ANN and GARCH) we conducted the results' analysis by comparing the results of the forecast with the historical data and by studying the forecast errors by the MSE, RMSE, MAE, Standard Deviation, the Theil's U and forecasting encompassing tests. It was found that the models developed by means of ANNs had lower MSE, RMSE and MAE than the GARCH (1,1) model and Theil U test indicated that the three models have smaller errors than those of a naïve forecast. Although the ANN based on returns have lower precision indicator values than those of ANN based on prices, the forecast encompassing test rejected the hypothesis that this model is better than that, indicating that the ANN models have a similar level of accuracy . It was concluded that for the data series studied the ANN models show a more appropriate Ibovespa forecasting than the traditional models of time series, represented by the GARCH model

Relevância:

100.00% 100.00%

Publicador:

Resumo:

In this paper artificial neural network (ANN) based on supervised and unsupervised algorithms were investigated for use in the study of rheological parameters of solid pharmaceutical excipients, in order to develop computational tools for manufacturing solid dosage forms. Among four supervised neural networks investigated, the best learning performance was achieved by a feedfoward multilayer perceptron whose architectures was composed by eight neurons in the input layer, sixteen neurons in the hidden layer and one neuron in the output layer. Learning and predictive performance relative to repose angle was poor while to Carr index and Hausner ratio (CI and HR, respectively) showed very good fitting capacity and learning, therefore HR and CI were considered suitable descriptors for the next stage of development of supervised ANNs. Clustering capacity was evaluated for five unsupervised strategies. Network based on purely unsupervised competitive strategies, classic "Winner-Take-All", "Frequency-Sensitive Competitive Learning" and "Rival-Penalize Competitive Learning" (WTA, FSCL and RPCL, respectively) were able to perform clustering from database, however this classification was very poor, showing severe classification errors by grouping data with conflicting properties into the same cluster or even the same neuron. On the other hand it could not be established what was the criteria adopted by the neural network for those clustering. Self-Organizing Maps (SOM) and Neural Gas (NG) networks showed better clustering capacity. Both have recognized the two major groupings of data corresponding to lactose (LAC) and cellulose (CEL). However, SOM showed some errors in classify data from minority excipients, magnesium stearate (EMG) , talc (TLC) and attapulgite (ATP). NG network in turn performed a very consistent classification of data and solve the misclassification of SOM, being the most appropriate network for classifying data of the study. The use of NG network in pharmaceutical technology was still unpublished. NG therefore has great potential for use in the development of software for use in automated classification systems of pharmaceutical powders and as a new tool for mining and clustering data in drug development

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Post dispatch analysis of signals obtained from digital disturbances registers provide important information to identify and classify disturbances in systems, looking for a more efficient management of the supply. In order to enhance the task of identifying and classifying the disturbances - providing an automatic assessment - techniques of digital signal processing can be helpful. The Wavelet Transform has become a very efficient tool for the analysis of voltage or current signals, obtained immediately after disturbance s occurrences in the network. This work presents a methodology based on the Discrete Wavelet Transform to implement this process. It uses a comparison between distribution curves of signals energy, with and without disturbance. This is done for different resolution levels of its decomposition in order to obtain descriptors that permit its classification, using artificial neural networks

Relevância:

100.00% 100.00%

Publicador:

Resumo:

The bidimensional periodic structures called frequency selective surfaces have been well investigated because of their filtering properties. Similar to the filters that work at the traditional radiofrequency band, such structures can behave as band-stop or pass-band filters, depending on the elements of the array (patch or aperture, respectively) and can be used for a variety of applications, such as: radomes, dichroic reflectors, waveguide filters, artificial magnetic conductors, microwave absorbers etc. To provide high-performance filtering properties at microwave bands, electromagnetic engineers have investigated various types of periodic structures: reconfigurable frequency selective screens, multilayered selective filters, as well as periodic arrays printed on anisotropic dielectric substrates and composed by fractal elements. In general, there is no closed form solution directly from a given desired frequency response to a corresponding device; thus, the analysis of its scattering characteristics requires the application of rigorous full-wave techniques. Besides that, due to the computational complexity of using a full-wave simulator to evaluate the frequency selective surface scattering variables, many electromagnetic engineers still use trial-and-error process until to achieve a given design criterion. As this procedure is very laborious and human dependent, optimization techniques are required to design practical periodic structures with desired filter specifications. Some authors have been employed neural networks and natural optimization algorithms, such as the genetic algorithms and the particle swarm optimization for the frequency selective surface design and optimization. This work has as objective the accomplishment of a rigorous study about the electromagnetic behavior of the periodic structures, enabling the design of efficient devices applied to microwave band. For this, artificial neural networks are used together with natural optimization techniques, allowing the accurate and efficient investigation of various types of frequency selective surfaces, in a simple and fast manner, becoming a powerful tool for the design and optimization of such structures

Relevância:

100.00% 100.00%

Publicador:

Resumo:

ln this work, it was deveIoped a parallel cooperative genetic algorithm with different evolution behaviors to train and to define architectures for MuItiIayer Perceptron neural networks. MuItiIayer Perceptron neural networks are very powerful tools and had their use extended vastIy due to their abiIity of providing great resuIts to a broad range of appIications. The combination of genetic algorithms and parallel processing can be very powerful when applied to the Iearning process of the neural network, as well as to the definition of its architecture since this procedure can be very slow, usually requiring a lot of computational time. AIso, research work combining and appIying evolutionary computation into the design of neural networks is very useful since most of the Iearning algorithms deveIoped to train neural networks only adjust their synaptic weights, not considering the design of the networks architecture. Furthermore, the use of cooperation in the genetic algorithm allows the interaction of different populations, avoiding local minima and helping in the search of a promising solution, acceIerating the evolutionary process. Finally, individuaIs and evolution behavior can be exclusive on each copy of the genetic algorithm running in each task enhancing the diversity of populations

Relevância:

100.00% 100.00%

Publicador:

Resumo:

Nowadays, where the market competition requires products with better quality and a constant search for cost savings and a better use of raw materials, the research for more efficient control strategies becomes vital. In Natural Gas Processin Units (NGPUs), as in the most chemical processes, the quality control is accomplished through their products composition. However, the chemical composition analysis has a long measurement time, even when performed by instruments such as gas chromatographs. This fact hinders the development of control strategies to provide a better process yield. The natural gas processing is one of the most important activities in the petroleum industry. The main economic product of a NGPU is the liquefied petroleum gas (LPG). The LPG is ideally composed by propane and butane, however, in practice, its composition has some contaminants, such as ethane and pentane. In this work is proposed an inferential system using neural networks to estimate the ethane and pentane mole fractions in LPG and the propane mole fraction in residual gas. The goal is to provide the values of these estimated variables in every minute using a single multilayer neural network, making it possibly to apply inferential control techniques in order to monitor the LPG quality and to reduce the propane loss in the process. To develop this work a NGPU was simulated in HYSYS R software, composed by two distillation collumns: deethanizer and debutanizer. The inference is performed through the process variables of the PID controllers present in the instrumentation of these columns. To reduce the complexity of the inferential neural network is used the statistical technique of principal component analysis to decrease the number of network inputs, thus forming a hybrid inferential system. It is also proposed in this work a simple strategy to correct the inferential system in real-time, based on measurements of the chromatographs which may exist in process under study

Relevância:

100.00% 100.00%

Publicador:

Resumo:

This work presents a diagnosis faults system (rotor, stator, and contamination) of three-phase induction motor through equivalent circuit parameters and using techniques patterns recognition. The technology fault diagnostics in engines are evolving and becoming increasingly important in the field of electrical machinery. The neural networks have the ability to classify non-linear relationships between signals through the patterns identification of signals related. It is carried out induction motor´s simulations through the program Matlab R & Simulink R , and produced some faults from modifications in the equivalent circuit parameters. A system is implemented with multiples classifying neural network two neural networks to receive these results and, after well-trained, to accomplish the identification of fault´s pattern

Relevância:

100.00% 100.00%

Publicador:

Resumo:

This work proposes hardware architecture, VHDL described, developed to embedded Artificial Neural Network (ANN), Multilayer Perceptron (MLP). The present work idealizes that, in this architecture, ANN applications could easily embed several different topologies of MLP network industrial field. The MLP topology in which the architecture can be configured is defined by a simple and specifically data input (instructions) that determines the layers and Perceptron quantity of the network. In order to set several MLP topologies, many components (datapath) and a controller were developed to execute these instructions. Thus, an user defines a group of previously known instructions which determine ANN characteristics. The system will guarantee the MLP execution through the neural processors (Perceptrons), the components of datapath and the controller that were developed. In other way, the biases and the weights must be static, the ANN that will be embedded must had been trained previously, in off-line way. The knowledge of system internal characteristics and the VHDL language by the user are not needed. The reconfigurable FPGA device was used to implement, simulate and test all the system, allowing application in several real daily problems