• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 139
  • 128
  • 75
  • 31
  • 15
  • 11
  • 6
  • 4
  • 3
  • 3
  • 2
  • 2
  • 2
  • 1
  • 1
  • Tagged with
  • 515
  • 515
  • 107
  • 97
  • 97
  • 78
  • 72
  • 71
  • 70
  • 66
  • 64
  • 60
  • 57
  • 50
  • 48
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
221

Μελέτη ηλεκτροεγκεφαλογραφήματος με βάση μεγέθη από τη θεωρία πληροφορίας

Ξενικού, Μόνικα Φιλίτσα 20 October 2010 (has links)
Στην παρούσα μεταπτυχιακή διπλωματική εργασία, η κυματιδιακή ανάλυση (Wavelet Analysis) εφαρμόζεται σε ηλεκτροεγκεφαλικές καταγραφές ασθενών που πάσχουν από επιληψία και σε προκλητά δυναμικά ασθενών που πάσχουν από σχιζοφρένεια και αλκοολισμό, με σκοπό την ανάδειξη παθολογικών προτύπων. Συγκεκριμένα, χρησιμοποιείται ο Διακριτός Κυματιδιακός Μετασχηματισμός (Discrete Wavelet Transform) για την εκτίμηση της Εντροπίας (Entropy) τωνηλεκτροεγκεφαλικών σημάτων. Η Εντροπία είναι ένα μέγεθος το οποίο εκφράζει το βαθμό τάξης ενός σήματος. Βιοσήματα με μεγάλο βαθμό οργάνωσης και περιορισμένο συχνοτικό περιεχόμενο -που θεωρείται ότι εκφράζουν πιο συντονισμένη εγκεφαλική λειτουργία- έχουν χαμηλές τιμές Εντροπίας. Αντίθετα, όσο πιο στοχαστικό είναι ένα σήμα τόσο μεγαλύτερη η τιμή της Εντροπίας του. Για τη μελέτη μας υλοποιήσαμε δύο διαφορετικά μεγέθη Εντροπίας, την Κυματιδιακή Εντροπία (Wavelet Entropy) και την Εντροπία Renyi (Renyi Entropy), τα οποία έχουν διαφορετικές σχέσεις ορισμού αλλά εκφράζουν και τα δύο το βαθμό τάξης των αναλυόμενων σημάτων. Τα αποτελέσματα της ανάλυσης αποκαλύπτουν ότι και τα δύο αυτά μεγέθη Εντροπίας κατορθώνουν να εντοπίσουν την επιληπτική κρίση, καθώς η τιμή τους διαφοροποιείται κατά τη διάρκειά της σε σχέση με την τιμή τους πριν από αυτή. Ακόμα, στατιστική ανάλυση των αποτελεσμάτων για προκλητά δυναμικά αποκαλύπτει ότι είναι δυνατός ο διαχωρισμός των παθολογικών καταγραφών (σχιζοφρενών ή αλκοολικών) από αυτές υγιών μαρτύρων, σε συγκεκριμένες περιοχές ενδιαφέροντος των σημάτων. / Wavelet Analysis of EEG during epileptic seizures and evoked potentials of patients suffering from schizophrenia and alcoholism was carried out in the present project. Discrete Wavelet Transform was used to estimate the Entropy of patients’ biosignals. Entropy is a measure of the degree of order of a signal -and subsequently of the system it represents- also reflecting the complexity of its power spectrum. An ordered activity, like a sinusoidal signal, is manifested as a narrow peak in the frequency domain which corresponds to a low entropy value. On the other extreme, a disordered activity (e.g. the one generated by pure noise or by a deterministic chaotic system) will have a wide band response, thus being reflected in higher entropies. In our research we used two differently defined Entropies, Wavelet Entropy and Renyi Entropy, both revealing the degree of order of the signal. Results show that both Entropy measures accomplish the goal they were used for. They manage to detect the epileptic seizure as their value clearly decreased during seizure compared to the pre-ictal value. Also, statistically significant differences were observed between entropy values of ERPs of healthy subjects and patients suffering from schizophrenia and alcoholism.
222

Manutenção preditiva de um par engrenado através da análise de lubrificantes e da análise de vibrações utilizando a transformada de wavelet / Predictive maintenance of a gearbox through lubricant analysis and vibration analysis using the wavelet transform

Pereira, André Luis Vinagre 27 February 2018 (has links)
Submitted by ANDRÉ LUIS VINAGRE PEREIRA null (andreluisvp@gmail.com) on 2018-03-29T18:17:43Z No. of bitstreams: 1 Dissertação Mestrado Final.pdf: 7001331 bytes, checksum: 858704904256f11c8131d5f17bd44a78 (MD5) / Approved for entry into archive by Cristina Alexandra de Godoy null (cristina@adm.feis.unesp.br) on 2018-04-02T12:53:22Z (GMT) No. of bitstreams: 1 pereira_alv_me_ilha.pdf: 7001331 bytes, checksum: 858704904256f11c8131d5f17bd44a78 (MD5) / Made available in DSpace on 2018-04-02T12:53:22Z (GMT). No. of bitstreams: 1 pereira_alv_me_ilha.pdf: 7001331 bytes, checksum: 858704904256f11c8131d5f17bd44a78 (MD5) Previous issue date: 2018-02-27 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Na manutenção preditiva, as análises dos sinais de vibração e das partículas do óleo são frequentemente utilizadas para o diagnóstico de falhas em redutores, porque elas contêm informações das condições de seus elementos mecânicos. Os sinais de vibração de um redutor geralmente têm muito ruído e a relação sinal-ruído é tão baixa que a extração de informações dos componentes do sinal é muito difícil, especialmente em situações práticas. Uma das soluções para este problema é a aplicação de técnicas de processamento do sinal para facilitar a obtenção de informações. Neste trabalho, uma técnica de cancelamento de ruído, a média temporal síncrona (TSA), e outra técnica da transformada contínua de wavelet de Morlet foram desenvolvidas para extração de recursos e diagnóstico de diferentes tipos de danos locais da engrenagem. Estas técnicas são aplicadas em sinais medidos em uma bancada experimental, que consiste em um par engrenado acoplado a um motor e a um gerador. Outro método para monitorar o estado do sistema é pela análise de partículas presente no óleo provenientes do desgaste das engrenagens. Avaliando a quantidade, formato, tamanho e material das partículas é possível obter informações das condições do equipamento e do tipo de desgaste ocorrido. Neste trabalho, foram feitas a análise do óleo pelas técnicas da ferrografia e contagem de partículas. A parte experimental deste trabalho foi dividida em dois experimentos. No primeiro experimento as condições de um par engrenado durante toda a sua vida útil foi monitorada, enquanto que no segundo experimento, um entalhe foi feito na raiz do dente simulando uma trinca por fadiga. A análise das partículas de óleo mostrou quais tipos de desgastes estava ocorrendo e a técnica da transformada contínua de wavelet mostrou-se precisa na identificação de falhas em dentes de engrenagens, sendo possível indicar em qual dente a falha estava se desenvolvendo. / At the predictive maintenance, the vibration signals analysis and oil particles analysis are frequently used to diagnose failures in a gearbox, because they contain information about the condition of its mechanic’s elements. The vibration signals of a gearbox usually have a lot of noise and the ratio ‘signal-noise’ is very low, making the extraction of information from the signals component very hard, especially in a practical situation. One of the solutions to this problem is the application of technics of signal processing, to improve the collection of information. At this study, a technique of noise cancellation, Temporal Synchronous Average (TSA) and another technique called continuous transform with the Morlet wavelet were executed for the extraction of resources and diagnostics of different type of gears local damages. Those methods are applied to signals measured on an experimental test stand, consisting of a gearbox with an engine and a generator. Another method for monitoring system wear is by analyzing wear particles in the oil generated due to the wear on the gears. By evaluating the quantity, shape, size and material of the particles it is possible to obtain information about the conditions of the equipment and the type of wear that has occurred. During this work, it was done the analysis of the oil by the techniques of ferrography and particle counting. The experimental part of this study was divided into two experiments. On the first experiment was monitored the conditions of a couple meshed throughout its useful life and in the second was made a notch in the root of the tooth simulating a crack by fatigue. The analysis of the oil particles showed what types of wear was occurring and the technique of the continuous wavelet transform was accurate in the identification of defects in gear's teeth, and it was possible to indicate which tooth was failing.
223

Manutenção preditiva de um par engrenado através da análise de lubrificantes e da análise de vibrações utilizando a transformada de wavelet /

Pereira, André Luis Vinagre. January 2018 (has links)
Orientador: Aparecido Carlos Gonçalves / Resumo: Na manutenção preditiva, as análises dos sinais de vibração e das partículas do óleo são frequentemente utilizadas para o diagnóstico de falhas em redutores, porque elas contêm informações das condições de seus elementos mecânicos. Os sinais de vibração de um redutor geralmente têm muito ruído e a relação sinal-ruído é tão baixa que a extração de informações dos componentes do sinal é muito difícil, especialmente em situações práticas. Uma das soluções para este problema é a aplicação de técnicas de processamento do sinal para facilitar a obtenção de informações. Neste trabalho, uma técnica de cancelamento de ruído, a média temporal síncrona (TSA), e outra técnica da transformada contínua de wavelet de Morlet foram desenvolvidas para extração de recursos e diagnóstico de diferentes tipos de danos locais da engrenagem. Estas técnicas são aplicadas em sinais medidos em uma bancada experimental, que consiste em um par engrenado acoplado a um motor e a um gerador. Outro método para monitorar o estado do sistema é pela análise de partículas presente no óleo provenientes do desgaste das engrenagens. Avaliando a quantidade, formato, tamanho e material das partículas é possível obter informações das condições do equipamento e do tipo de desgaste ocorrido. Neste trabalho, foram feitas a análise do óleo pelas técnicas da ferrografia e contagem de partículas. A parte experimental deste trabalho foi dividida em dois experimentos. No primeiro experimento as condições de um par engrenado d... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: At the predictive maintenance, the vibration signals analysis and oil particles analysis are frequently used to diagnose failures in a gearbox, because they contain information about the condition of its mechanic’s elements. The vibration signals of a gearbox usually have a lot of noise and the ratio ‘signal-noise’ is very low, making the extraction of information from the signals component very hard, especially in a practical situation. One of the solutions to this problem is the application of technics of signal processing, to improve the collection of information. At this study, a technique of noise cancellation, Temporal Synchronous Average (TSA) and another technique called continuous transform with the Morlet wavelet were executed for the extraction of resources and diagnostics of different type of gears local damages. Those methods are applied to signals measured on an experimental test stand, consisting of a gearbox with an engine and a generator. Another method for monitoring system wear is by analyzing wear particles in the oil generated due to the wear on the gears. By evaluating the quantity, shape, size and material of the particles it is possible to obtain information about the conditions of the equipment and the type of wear that has occurred. During this work, it was done the analysis of the oil by the techniques of ferrography and particle counting. The experimental part of this study was divided into two experiments. On the first experiment was monitored the... (Complete abstract click electronic access below) / Mestre
224

Simulation, measurement and detection of leakage and blockage in fluid pipeline systems

Owowo, Julius January 2016 (has links)
Leakage and blockage of oil and gas pipeline systems, water pipelines, pipe-work of process plants and other pipe networks can cause serious environmental, health and economic problems. There are a number of non-destructive testing (NDT) methods for detecting these defects in pipeline systems such as radiographic, ultrasonic, magnetic particle inspection, pressure transient and acoustic wave methods. In this study, the acoustic wave method and a modal frequency technique, which based on a roving mass method, are used. The aim of the thesis is to employ acoustic wave propagation based methods in conjunction with stationary wavelet transform (SWT) to identify leakage and blockage in pipe systems. Moreover, the research is also aimed at using the difference of modal frequencies of fluid-filled pipes with and without defects and a roving mass, and consequently, to develop a roving mass-based defect detection method for pipelines. In the study, the acoustic finite-element analysis (AFEA) method is employed to simulate acoustic wave propagation in small and large air-filled water-filled straight pipe and pipe networks with leakage and blockage but without flow. Computational fluid dynamics (CFD) analysis was also employed to simulate acoustic wave propagation in air-and water-filled pipes with flow, leakage and blockage. In addition, AFEA was used to predict the modal frequencies of air- and water-filled pipes with leakage and blockage in the presence of a roving mass that was traversed along the axis of the pipes. Experimental testing was conducted to validate some of the numerical results. Two major experiments were performed. The first set of experiments consisted of the measurement of acoustic wave propagation in a straight air-filled pipe with leakage and blockage. The second set of experiments concerned the measurement of acoustic wave propagation in an air-filled pipe network comprising straight pipe, elbows and T-piece and flange. The AFEA and CFD analysis of fluid-filled pipe can be used to simulate the acoustic wave propagation and acoustic wave reflectometry of a fluid-filled pipe with leakage and blockage of different sizes down to a small leakage size of 1mm diameter and a blockage depth of 1.2mm in a pipe. Similarly, the AFEA method of a static fluid-filled pipe can be used to simulate acoustic wave modal frequency analysis of a fluid-filled pipe with leakage and blockage of different sizes down to a leakage of 1mm diameter and a blockage depth of 1.2mm. Moreover, the measured signal of acoustic wave propagation in an air-filled can be successfully decomposed and de-noised using the SWT method to identify and locate leakages of different sizes down to 5mm diameter, and small blockage depth of 1.2mm. Also, the SWT approximation coefficient, detail and de-noised detail coefficient curves of an air-filled pipe with leakage and blockage and a roving mass give leakage and blockage indications that can be used to identify, locate and estimate the size of leakage and blockage in a pipe.
225

Um método de avaliação da amplitude do potencial P300 comparando indivíduos com alto risco e baixo risco para o alcoolismo

Lopes, Carla Diniz January 2010 (has links)
A ocorrência de variações nos sinais de eletroencefalograma (EEG) de indivíduos que apresentam predisposição a desenvolver a doença do alcoolismo é conhecida e documentada na literatura médica e científica. Dentre as possíveis variações, encontram-se as anormalidades no potencial relacionado ao evento (ERP) P300, um dos principais endofenótipos da doença. Geralmente, este componente tem uma amplitude significativamente menor em indivíduos com alto risco (AR) de desenvolver a doença, quando comparada à amplitude observada em sinais de indivíduos com baixo risco (BR). A técnica atualmente empregada para distinguir os sinais de ERPs P300 dos indivíduos com AR e BR para desenvolver o alcoolismo é baseada na análise visual da amplitude máxima no domínio do tempo e do espectro de frequencias do sinal, obtido através da transformada de Fourier. O objetivo deste trabalho é contribuir para o estudo da identificação da predisposição ao alcoolismo, utilizando técnicas de processamento de sinais, como a transformada wavelet (WT), e de inteligência artificial, por meio das redes neurais artificiais (ANNs). A WT foi utilizada por ser mais adequada ao tratamento de sinais como os ERPs (sinais nãoestacionários), quando comparada, por exemplo, à transformada de Fourier. As redes neurais possibilitam a automatização do processo de identificação dos diferentes grupos. Através de um sistema híbrido formado por estas duas técnicas, pretende-se extrair características de sinais de ERP que identifiquem indivíduos com predisposição ao alcoolismo, e automatizar a identificação destes indivíduos. No desenvolvimento da pesquisa, foi identificada a necessidade de aplicar um préprocessamento aos sinais de ERP, preparando-os para a transformação wavelet. Os coeficientes wavelet assim obtidos formaram os dados de entrada que alimentaram as (ANNs), as quais utilizaram o algoritmo de erro backpropagation no treinamento. Com as técnicas utilizadas, após o treinamento, as ANNs foram capazes de classificar cerca de 90% dos sinais de ERP dos indivíduos com AR e BR. / The occurrence of variations in electroencephalogram (EEG) signals of individuals who are predisposed to develop the disease of alcoholism is known and documented in the medical and scientific literature. Among these variations, are the abnormalities in the event related potential (ERP) P300, a major endophenotype of this disease. Generally, this component has an amplitude significantly smaller in patients at high risk (HR) of developing the disease when compared to the amplitude seen in the signals of individuals with low risk (LR). The technique currently used to distinguish signals of P300 ERPs in individuals with HR and LR for developing alcoholism is based on visual analysis of the maximum amplitude in the time domain and of the frequency spectrum of the signal, obtained via Fourier transform. The aim of this thesis is to study the identification of predisposition to alcoholism, by techniques of signal processing such as wavelet transform (WT) and artificial intelligence through artificial neural networks (ANNs). The WT was used because it is more appropriate for processing signals such as ERP (non-stationary signals), when compared, for example, to the Fourier transform. Neural networks enable the automation of the process of identifying the different groups. Using a hybrid system formed by these two techniques, it is intended to extract features of ERP signals that identify individuals predisposed to alcoholism, and automate the identification of these individuals. The research has identified the need to apply a pre-processing to the signals of ERP, preparing them for the wavelet transformation. The wavelet coefficients thus obtained formed the input data to fed the ANNs, which used the error algorithm backpropagation in training. Using these techniques, after training, the ANNs were able to classify about 90% of ERP signs of individuals with LR and HR.
226

Proteção diferencial de transformadores trifásicos utilizando a transformada wavelet

Oliveira, Mario Orlando January 2009 (has links)
A qualidade e a continuidade do fornecimento de energia elétrica aos consumidores são fatores muito importantes quando da avaliação da eficiência de um sistema elétrico de potência. Nesse contexto, os transformadores são equipamentos muito importantes e demandam especial atenção quando do projeto do esquema de proteção. Apesar do crescente desenvolvimento das metodologias de proteção de transformadores trifásicos, alguns aspectos ainda não foram totalmente solucionados. Um desses diz respeito à proteção diferencial de transformadores de potência, a qual apresenta vários problemas na discriminação de faltas internas ao transformador. A geração de correntes diferenciais provocada por fenômenos transitórios, como a energização do transformador, produz a incorreta operação do relé, ocasionando uma queda na eficiência do esquema de proteção diferencial. Assim sendo, o presente trabalho apresenta uma nova metodologia de proteção diferencial de transformadores trifásicos, a qual utiliza a transformada wavelet para extrair os sinais transitórios dominantes induzidos pelas faltas internas. A transformada wavelet é uma eficiente ferramenta utilizada no estudo de sinais não-estacionários e de rápida transição. De forma a atender os principais problemas do esquema convencional de proteção, a transformada wavelet discreta é utilizada para decompor os sinais de corrente diferencial em várias faixas de freqüências. Após essa decomposição, a variação de energia espectral dos coeficientes de detalhe wavelet é analisada pelo algoritmo proposto, e assim uma discriminação entre faltas internas e correntes de magnetização, ou correntes inrush, é feita. Usando um modelo elaborado de um sistema elétrico de transmissão são efetuadas rigorosas simulações computacionais para avaliar o desempenho do algoritmo proposto. Os resultados obtidos nessas simulações mostram que a metodologia de proteção diferencial de transformadores trifásicos baseada na variação de energia espectral dos coeficientes wavelets apresenta um ótimo desempenho quando comparada com a metodologia de proteção convencional. / Power supply quality and continuity are very important aspect when assessing the efficiency of an electric power system. In this context, the transformers are key equipments that require special attention during the protection scheme design. Despite the increasing development of methodologies for three-phase transformers protection, some aspects have not yet been fully studied. One of these aspects concerns to the differential protection of power transformers, which presents several restrictions regarding the characterization of internal faults. The observation of differential currents caused by transient phenomena such as transformer energization, produces an incorrect operation of protective relaying, causing a drop in the protection scheme efficiency. Therefore, this work presents a new methodology for differential protection of three-phase transformers using the wavelet transform to extract the transient signals induced by the dominant internal faults. The wavelet transform is an efficient tool in the study of non-stationary signals with fast transients. In order to overcome the main problems of the traditional protection scheme, the discrete wavelet transform is used to decompose the differential current signals into several bands of frequencies. After this decomposition, the spectral energy variation of the wavelet detail coefficients is analyzed by the proposed algorithm and, thus, classification between internal faults, external faults and inrush currents is performed. Using a transmission system model, accurate simulations are performed to evaluate the computational performance of the proposed protection algorithm. The results obtained in these simulations show that the proposed methodology has a great performance when compared with traditional protection philosophies.
227

Análise de Sinais Pulmonares Utilizando Técnicas no Domínio Tempo-Frequência e Classificação Neural. / Signal Analysis in Lung Using Techniques Time-Frequenci Domain and Neural Classification.

Almeida, Alberto Jorge Santos de 25 November 2010 (has links)
Auscultation is a method of clinical practice, simple, noninvasive, used to diagnose diseases of the respiratory system. However, it is an imprecise method because, among other factors, the limitations of the auditory system, the overlap of heart sounds and human hearing sensitivity difference, besides the limited spectral response characteristic of many commercial stethoscopes. These factors contribute to the diagnosis relies heavily on the experience of the professional expert. The acoustic analysis of spectral characteristics of signals of ventilation can be a complementary diagnostic technique in facilitating the process of detection and identification of breath sounds, providing aid in the assessment of symptom severity and treatment efficacy. In this study, we attempted to structure a process of analysis of pulmonary signs to identify characteristics of respiratory disorders. Accordingly, the signals were processed by filtering processes and decomposed into sub-frequency bands through discrete wavelet transform (DWT), generating vectors as coefficients for classification using an Artificial Neural Network. A case study with signals obtained from tests was presented and duly considered. / A ausculta é um método da prática clinica, simples, não invasivo, utilizado no diagnóstico de doenças do sistema respiratório. Porém, trata-se de um método impreciso devido, entre outros fatores, às limitações do sistema auditivo, a sobreposição de sons cardíacos e a diferença de sensibilidade auditiva humana, além da característica de resposta espectral limitada de muitos estetoscópios comerciais. Tais fatores contribuem para que o diagnóstico dependa muito da experiência do profissional especialista. A análise acústica de características espectrais de sinais da ventilação pulmonar pode ser uma técnica complementar de diagnóstico, facilitando o processo de detecção e identificação desses sons respiratórios, possibilitando auxiliar a avaliação da gravidade dos sintomas e a eficácia do tratamento. Neste trabalho, buscou-se estruturar um processo de análise de sinais pulmonares, para a identificação de características das patologias respiratórias. Nesse sentido, os sinais foram tratados por processos de filtragem e decompostos em sub-bandas de freqüências através da Transformada de Wavelet Discreta (DWT), gerando vetores como coeficientes para classificação utilizando uma Rede Neural Artificial. Um estudo de caso com sinais obtidos por testes foi apresentado e devidamente analisado.
228

Detecção das interações do sistema brisa marinha/terrestre com sistemas sinóticos na costa leste de Alagoas utilizando Transformada Wavelet. / Detection of the interactions of the sea/land breeze systems with synoptic systems on the eastern coast of Alagoas using wavelet transform technique.

Ferreira, André Deodato 11 December 2009 (has links)
The eastern coast of the Northeast Brazil (NEB) is a region where transient weather systems interact with local systems, modulating their frequency and intensity. Understanding the mechanisms of such interactions may be instrumental in explaining the observed precipitation patterns in this region. The objective of this work is quite within this scope and takes a new approach, namely the study of time series using spectral decomposition. The wavelet transform was adopted for it has proven to be a very powerful tool in dealing with local data in many areas. It furnishes a hierarchical frame that allows the double location in time and frequency domains. The hourly averaged series of wind speed and direction, air temperature, humidity and precipitation were taken at an observing tower at the 12.40 high during the October, 2004 October 2005 period in the Santa Rita island (mangrove area) at 9° 42 18 S e 35° 48 32 W. Relevant characteristics of the seasonality, transport and intensity of the breeze systems were noted, showing a clear distinction among the rainy, dry and transitional periods. Mesoscale systems are more active during the dry period because they are quite dependent of the daily cycle of heating and cooling. Large scale transient systems are more prominent during the rainy season, being responsible for the highest observed precipitation rates. The local wind system of sea land breezes shows a predominant southeast (SE) direction, thus evidencing the sea breeze. It was also noted that the highest amplitudes and wind persistency from SE were reinforced by the SE Trade Winds, creating a constructive interference in the signal. The prevailing winds from northwest (NW) were only observed during the dawn, thus characterizing the land breeze branch of the local circulation. The sea/land breeze variability affected the precipitation regime, by forming a surface convergence zone (rising motions) in accord with the time of the precipitations. These also showed a conic multifractal pattern in the periodgram, typical of high space-time variability. / Fundação de Amparo a Pesquisa do Estado de Alagoas / A costa leste do NEB é palco da ação simultânea de fenômenos transientes que interagem com os sistemas locais, modificando sua freqüência e intensidade. O conhecimento destas interações pode revelar o padrão das precipitações junto à costa do NEB. O objetivo desta dissertação se encaixa neste contexto e aborda os dados sob uma perspectiva nova, que consiste no estudo das séries temporais através de decomposição espectral. Para este fim, adotou-se a Transformada em Ondeletas (TO), que nas ultimas décadas tem alcançado notável êxito em diversas áreas da ciência. Trata-se de uma ferramenta matemática poderosa na detecção das freqüências e intensidades das séries temporais. Esta técnica fornece uma moldura hierárquica que permite a dupla localização em tempo e freqüência. Os dados horários de velocidade e direção do vento, temperatura do ar, umidade e precipitação foram medidos numa torre ao nível de 12,40m acima do solo, durante o período (Outubro de 2004 Outubro de 2005) numa área de proteção ambiental da Ilha de Santa Rita, localizada no mangue natural a 9° 42 18 S e 35° 48 32 W. Foram patenteadas características relevantes de transporte, sazonalidade e intensidade dos sistemas de brisas. As características do sinal mostraram a distinção entre as quadras seca e chuvosa, bem como as respectivas transições. Os sistemas de mesoescala atuam com mais intensidade no período seco, pelo fato de se ajustar ao ciclo diário da forçante térmica (radiação solar). Para a quadra chuvosa, predominam os sistemas transientes de grande escala, que são os responsáveis pelos maiores índices pluviométricos. O regime de ventos locais, caracterizado pelos sistemas das brisas de terra e mar, assinala que a direção do vento predominante é de sudeste (SE), o que caracteriza a brisa marítima. Foi verificado que as maiores amplitudes e persistência dos ventos de SE foram incrementadas pelos ventos Alísios de SE, que sopraram o ano inteiro, criando uma interferência construtiva no sinal. Os ventos predominantes de noroeste (NW) só foram observados no período da madrugada, distintivos da brisa terrestre. A mudança das brisas terrestres/marítimas apresentou intervenção no sistema de precipitação, criando uma zona de convergência em superfície (movimentos ascendentes) corroborando com os horários em que ocorreram as precipitações. Estas exibiram uma configuração multifractal cônica no periodograma, típica da sua alta variabilidade espaço-temporal.
229

IMAGE-BASED MODELING AND PREDICTION OF NON-STATIONARY GROUND MOTIONS

DAK HAZIRBABA, YILDIZ 01 May 2015 (has links)
Nonlinear dynamic analysis is a required step in seismic performance evaluation of many structures. Performing such an analysis requires input ground motions, which are often obtained through simulations, due to the lack of sufficient records representing a given scenario. As seismic ground motions are characterized by time-varying amplitude and frequency content, and the response of nonlinear structures is sensitive to the temporal variations in the seismic energy input, ground motion non-stationarities should be taken into account in simulations. This paper describes a nonparametric approach for modeling and prediction of non-stationary ground motions. Using Relevance Vector Machines, a regression model which takes as input a set of seismic predictors, and produces as output the expected evolutionary power spectral density, conditioned on the predictors. A demonstrative example is presented, where recorded and predicted ground motions are compared in time, frequency, and time-frequency domains. Analysis results indicate reasonable match between the recorded and predicted quantities.
230

[en] IMAGE TRANSMISSION THROUGH NOISY CHANNELS WITH LT CODES / [pt] TRANSMISSÃO DE IMAGEM ATRAVÉS DE CANAL RUIDOSO USANDO CÓDIGOS LT

CARLOS MARIO CORREA TORRES 13 July 2010 (has links)
[pt] Para transmissão da informação de maneira confiável, em canais com apagamento, foram criados os códigos LT (Luby Transform), uma das principais classes de códigos fontanais. Estes códigos não têm uma taxa fixa, em outras palavras, eles têm taxa versátil. Esta dissertação aborda o estudo da transmissão de imagens através de canal ruidoso, AWGN (Aditive White Gaussian Noise), com o uso de Códigos LT. Investigou-se o desempenho usando uma modulação BPSK, dois esquemas foram testados: Um esquema para canal que inclui apagamento (BESC) e um outro que foi proposto usando um código Hamming em série com um código LT. O esquema LT-Hamming apresentou um ganho de código maior que o esquema BESC e o código convolucional de semelhantes características. Foi testado o esquema LT-Hamming para diferentes tipos de imagens em um canal AWGN usando a técnica SPIHT para a compressão das imagens. Para obter uma medida objetiva da qualidade da imagem recuperada foi usado o parâmetro PSNR (Peak Sinal to Noise Ratio) e foram apresentadas algumas imagens com o objetivo de analisar sua qualidade através de uma inspeção visual. Dado que o código LT é versátil para o que diz respeito à taxa de código, foi proposto um método para método para atribuir diferentes níveis de proteção da informação codificada, UEP (Unequal Error Protection). / [en] To transfer reliably information in erasure channels, LT (Luby Transform) codes were created, they are part of the main class of fountain codes, this codes don’t have fixed rate, in other words, they have a versatile code rate. This thesis address to the study of images transmission through noisy channel, AWGN (Aditive White Gaussian Noise) using LT codes. We investigated the performance using a BPSK modulation, two schemes were tested: A scheme of channel that includes deletion (BESC) and another that was proposed, using a Hamming code in series with a LT code. The LT-Hamming scheme present a gain code larger than BESC scheme and convolutional codes of similar characteristics. Was tested LT-Hamming scheme for different types of images on AWGN channel using the SPIHT technique for images compression. To obtain an objective measure of image quality was used the PSNR (Peak Signal Noise Ratio) and some images were presented in order to analize its quality through visual inspection given that LT code is a versatile for what concern the code rate it was proposed a method to assign different protection levels to the code information, UEP (Unequal Error Protection).

Page generated in 0.1858 seconds