Os algoritmos para síntese de circuitos digitais em geral visam a melhoria de uma função de custo composta de quatro critérios: área, desempenho, potência e testabilidade. Normalmente estes algoritmos conseguem uma relação de compromisso para a otimização de dois critérios. Efeitos indesejáveis também podem surgir com a otimização de um destes critérios. Por exemplo, as otimizações de desempenho podem introduzir falhas de colagem não testáveis (redundâncias) em um circuito, reduzindo a sua testabilidade. Muitos algoritmos de síntese lógica exploram propriedades específicas de determinadas funções a serem sintetizadas. Um exemplo de função com propriedades específicas são as funções ditas unate. Um exemplo deste tipo de função é o sinal de carry de um somador completo. Este tipo de função exige cuidados especiais para evitar a introdução de redundâncias. Muitos dos algoritmos para síntese lógica empregam a decomposição de Shannon para melhorar o desempenho de um circuito. A equação geral da decomposição de Shannon é expressa através de uma função binate. As redundâncias sempre serão introduzidas nos circuitos quando uma equação binate é utilizada para representar uma função unate. Diagramas de Decisão Binária (BDDs) são um tipo estruturas de dados muito utilizadas em algoritmos para síntese lógica. A decomposição de Shannon também é utilizada para derivar circuitos a partir de BDDs. Este tipo de estrutura representa uma função lógica, mas não mantém uma representação sem redundâncias da mesma. Infelizmente, os circuitos derivados a partir desta estrutura poderão ser redundantes, principalmente quando a decomposição de Shannon for utilizada. Existem estruturas de dados capazes de representar uma função sem redundâncias. Este é o caso dos VPBDDs , que possuem propriedades especiais que preservam características de testabilidade da função representada. Baseando-se nas propriedades dos VPBDDs, um novo algoritmo para remoção de redundâncias foi proposto. Este algoritmo é capaz de gerar circuitos sem redundâncias, mesmo quando a função, que é representada pelo VPBDD, é unate. Além da geração de circuitos sem redundâncias, o algoritmo garante que o atraso do circuito não aumenta após a remoção de redundâncias. A área dos circuitos resultantes pode aumentar, diminuir ou permanecer a mesma, considerando o número de portas lógicas utilizadas. Todos os resultados obtidos neste trabalho mostram que o algoritmo consegue realizar a remoção de redundâncias, sem prejudicar o atraso do circuito. Além disso, todos os caminhos redundantes do circuito têm seu atraso reduzido, pois com a remoção de redundâncias o número de portas lógicas em série é reduzido. A aplicação deste algoritmo apresenta bons resultados para circuitos aritméticos. Isto se deve principalmente ao fato do carry ser uma função unate, o que pode introduzir redundâncias no circuito se esta propriedade (de ser unate) não for tratada adequadamente. O algoritmo proposto também abre possibilidades para a criação de outras ferramentas de CAD, como por exemplo: uma ferramenta para análise de timing, um gerador de circuitos aritméticos sem redundâncias, ou ainda uma ferramenta para geração de teste, incluindo lista de falhas, vetores de teste e cobertura de falhas. / Algorithms for digital circuit design aim the reduction of a cost function composed of four criteria: area, delay, power and testability. Usually these algorithms are able to obtain a trade-off for the optimization of two of these criteria. Undesired effects may occur due to the optimization of one of the criteria. For instance, delay optimizations may introduce non testable stuck-at faults (redundancies) in a circuit, this way reducing its testability. Several logic synthesis algorithms exploit specific properties of the logic functions to be synthesized. One example of function with specific properties are the socalled unate functions. An example of this kind of function is the carry-out sign in a full adder circuit. This kind of function require special care in order to avoid redundancy introduction. Shannon decomposition [SHA 38] is used in many logic synthesis algorithms for improving circuit performance. The general case of the Shannon decomposition is represented by a binate (not unate) equation. Redundancies are introduced in a circuit when a binate equation is used to express a unate function. Binary Decision Diagrams (BDDs) are a kind of data structures widely used in the field of logic synthesis. Shannon decomposition is also used to derive circuits from BDDs. This data structure is used to represent logic functions, but it is not able to maintain an irredundant representation of any logic function. Unfortunately, circuits derived from BDDs will possibly have redundancies, specially when Shannon decomposition is used. Some data structures are able to represent any logic function in a irredundant form. This is the case of the VPBDDs [REI 95a] [REI 2000], which have special properties that preserve the testability properties of the functions being represented. Based on VPBDD properties, a novel algorithm for redundancy removal was proposed [MAR 2002]. This algorithm is able to generate irredundant circuits even when the function represented by the VPBDD is unate. In addition to the generation of irredundant circuits, the algorithm guarantees that the circuit delay will not be increased by redundancy removal. The final area may be increased, reduced or even remain the same, considering the number of logic gates. The results obtained in this work indicate that the algorithm is able to perform redundancy removal without increasing the circuit delay. Besides, all the redundant paths in the circuit have their delay reduced, as the number of logic gates in series will be reduced by the redundancy removal process. The application of this algorithm gives good results for arithmetic circuits. This is mainly due to the fact that the carry chain is composed of unate functions, this way redundancies are introduced in the circuit if this property is not adequately treated. The proposed algorithm allows for the creation of other CAD tools, as for instance: a timing analysis tool, a generator of irredundant arithmetic circuits, or even a test generation tool, including list of faults, test vectors as well as fault coverage.

Submitted by Mario BC (mario@bc.ufrpe.br) on 2016-08-10T14:29:56Z No. of bitstreams: 1 Flavia Rosa Santoro.pdf: 1115680 bytes, checksum: 425c8ab07e5a860cac3f849ed621f878 (MD5) / Made available in DSpace on 2016-08-10T14:29:56Z (GMT). No. of bitstreams: 1 Flavia Rosa Santoro.pdf: 1115680 bytes, checksum: 425c8ab07e5a860cac3f849ed621f878 (MD5) Previous issue date: 2014-02-21 / The Utilitarian Redundancy Model (URM) infers about socioecological systems resilience based on an ecological model. According to this model, the redundancy of species with the same utilitarian function ensures systems resilience. Under this perspective, the study about resilience of local medical systems (LMS) allows one to detect essential matters in order to understand how local communities can hold themselves and evolve in specific environments, especially where the access to natural resources and public health is limited. This dissertation aims to investigate the resilience of two LMS in Caatinga and to test some URM predictions. Therefore, our questions are based in features that give LMS resilience which follow the model's predictions: Do medicinal categories have redundancy prevalence? In the same medicinal category, are the redundant species used in the absence of preferred species? Is redundancy bigger in more severe and/or more frequent illnesses? Is information about illness treatment shared? The results show that the medical systems are highly specialized. However, when there is redundancy, the redundant species are the main strategy used when a disturbance occurs. This confirms an URM assumption in which redundancy ensures the maintenance of therapeutic functions. The redundancy of an illness was founded to be related to its frequency of occurrence, what guarantees the maintenance of medicinal functions in usual events. Despite this fact, the low rates of sharing we have found and the little redundancy in severe illnesses evidence points of vulnerability in the system. Based in our data, we might suggest that the studied LMS are mainly structured by means of specialists individual experiences and that the addition of new plants follows the illness dynamics of occurrence and considers life-threatening in the treatment. Under this perspective, the low sharing and the little redundancy in serious illnesses might reflect a system's evolutionary process feature. / O Modelo de Redundância Utilitária (MRU) se inspira em um modelo ecológico para inferir sobre a resiliência de sistemas socioecológicos. Segundo o modelo, a redundância de espécies com mesma função utilitária assegura a resiliência de um sistema. Nessa perspectiva, a análise da resiliência de sistemas médicos locais (SML) pode revelar pontos essenciais para compreender como comunidades locais conseguem se manter em determinados ambientes, principalmente onde o acesso a saúde pública e a recursos naturais é limitado. Esta dissertação visa investigar a resiliência de dois SMLs inseridos na Caatinga e testar algumas predições do MRU. Dessa forma as perguntas levantadas se baseiam em características que conferem resiliência ao SML a partir das premissas do modelo: Há predomínio de redundância nas categorias medicinais? Espécies redundantes são utilizadas na ausência das espécies preferidas da mesma categoria medicinal? Há maior redundância em enfermidades mais graves e/ou mais frequentes? Há compartilhamento das informações sobre o tratamento das enfermidades? Os sistemas médicos se mostraram altamente especializados, no entanto, quando há redundância, as espécies redundantes são empregadas na ocasião de um distúrbio, confirmando uma previsão do MRU em que a redundância garante a manutenção das funções terapêuticas. Foi encontrado que a redundância de uma enfermidade está associada com a sua freqüência de ocorrência, o que assegura a manutenção das funções medicinais em eventos recorrentes. No entanto, o baixo compartilhamento encontrado e a pouca redundância de plantas em enfermidades graves mostram pontos de vulnerabilidade do sistema. Com base em nossos dados, podemos sugerir que os SMLs estudados são estruturados principalmente através de experiências individuais dos especialistas, e que o acréscimo de novas plantas acompanha a dinâmica de ocorrência das enfermidades e considera o risco de vida no tratamento. Sob essa perspectiva, o baixo compartilhamento e a pouca redundância em enfermidades graves podem refletir uma característica do processo evolutivo que o sistema se encontra.

Made available in DSpace on 2016-08-17T14:53:09Z (GMT). No. of bitstreams: 1 Enio Aguiar de Araujo.pdf: 4163087 bytes, checksum: 164f9ddea4dce2f8f19d8e66d5e8294f (MD5) Previous issue date: 2010-05-28 / Typically, in the digital processing of electrocardiography signal, linear transformations are used to turn the signals more tractable in accordance to some application. For applications such as classification or data compression, it usually aimed to reduce the redundancy present in the signals, increasing the potential of the applications. There are various methods usually used for the task, the Fourier transform, the wavelet transform and principal component analysis. All those methods have any sort of limitation, being the use of a predefined space, orthogonal spaces or the limitations to second order statistics. In this work we propose the use of the independent component analysis method for the encoding of the ECG signals, using as theoretical basis the neuroscience concept of efficient coding. Two important results were found, the basis functions space generated by the proposed method is different from the spaces seen on the usual methods, and, on average, the method can reduce the redundancy of the signal. We concluded that the traditional methods might not exploit the coding potential of ECG signals due to their limitations, and also that ICA might be a reliable method for improving the performance comparing to the current systems. / Tipicamente, em processamento digital de sinais de eletrocardiografia, são utilizadas transformações lineares visando tornar os sinais mais tratáveis de acordo com a aplicação. Para aplicações como classificação ou compressão de dados, normalmente temse como objetivo dessas transformações, reduzir a redundãncia presente nesses sinais, o que aumenta o potencial dessas aplicações. Existem diversos métodos usualmente utilizados para essa tarefa, como a transformada de Fourier, a transformada de Wavelets, e análise de componentes principais. Todos esses métodos tem alguma limitação, seja a utilização de um espaço predefinido e ortogonal ou utilizar apenas estatística de segunda ordem. Neste trabalho propomos a utilização do método de análise de componentes independentes para a codificação de sinais de ECG utilizando como base o conceito da neurociência de codificação eficiente. Dois resultados importantes foram obtidos, o espaço de funções bases gerado pelo método proposto se diferencia dos espaços de transformações utilizados usualmente, e, em média, o método proposto teve maior capacidade de reduzir a redundância dos sinais. Concluímos que os métodos tradicionais podem não explorar ao máximo o potencial de codificação de sinais de ECG devido às suas limitações, e que ICA pode ser um método plausível para melhorarmos os sistemas atualmente utilizados.

Arabidopsis thaliana, s’est adaptée à la variété des niveaux de Pi dans le sol en développant 9transporteurs de phosphates, membres de la famille PHT1, intervenant dans l’acquisition de cet ionpar les racines et sa translocation. Ces protéines révèlent une très forte homologie de séquence entre elles (plus que 61%). La présence de certains transporteurs de type PHT1 dans différents types d’organes ainsi que le chevauchement fréquent entre les divers membres de la famille PHT1 témoigne de la complexité de leurs rôles. De plus, leur redondance génétique et fonctionnelle empêche l’analyse de leur rôle spécifique. En vue de s’affranchir de ces obstacles, notre approche combine plusieurs stratégies génétiques avec l’insertion d’une construction RNAi inactivant plusieurs membres de la famille PHT1 et en particulier le cluster localisé sur le chromosome 5 (PHT1;1/1;2/1;3). Ces outils génétiques ont révélé aussi le fonctionnement des protéines PHT1 à la fois en tant que transporteurs à basse et à haute affinité, ce qui suggère que leur activité est contrôlée au niveau post-traductionnel. En cas de carence en Pi, ces lignées affichent des modifications physiologiques (biomasse, rendement,…) dues à une forte réduction affectant l’activité de l’influx en phosphate (80 à 96%). Ce travail suggère que la redondance génétique et les mécanismes de compensations pourraient protéger la plante de l’inactivation de PHT1. Il a aussi révélé que la perception systémique du Pi est déclenchée par des mécanismes en aval de l’activité des PHT1. / Arabidopsis thaliana absorb inorganic phosphate (Pi) from the soil through an active transport process mediated by the 9 members of the PHT1 family. These proteins share a high level of similarity (greater than 61%), with overlapping expression patterns. The resulting genetic and functional redundancy prevents the analysis of their specific roles. To overcome this difficulty, our approach combined several mutations with gene silencing to inactivate multiple members of the PHT1 family, including a cluster of genes localized on chromosome 5 (PHT1;1, PHT1;2 and PHT1;3). Physiological analyses of these lines established that these three genes, along with PHT1;4, are the main contributors to Pi uptake. Furthermore, PHT1;1 plays an important role in translocation from roots to leaves in high phosphate conditions. These genetic tools also revealed that some PHT1 transporters likely exhibit a dual affinity for phosphate, suggesting that their activity is posttranslationally controlled. These lines display significant phosphate deficiency-related phenotypes (e.g. biomass and yield) due to a massive (80 to 96%) reduction in phosphate uptake activities. These defects limited the amount of internal Pi pool, inducing compensatory mechanisms triggered by the systemic Pi starvation response. Such reactions have been uncoupled from PHT1 activity suggesting that systemic Pi sensing is most probably acting downstream of PHT1.

Le travail mené dans cette thèse est une contribution au développement mécatronique d'interfaces haptiques pour un système de télé-opération dédié aux applications médicales du type chirurgie mini-invasive. Dans un premier temps, nous avons mené une évaluation d'une interface « maître » existante, ayant une architecture parallèle sphérique et développée au sein de l'équipe robotique de l'Institut PPRIME. Cette évaluation a montré la présence de singularités, en particulier des singularités parallèles, à l'intérieur de l'espace du travail de l'interface. La présence de singularités altère le comportement cinématique en amplifiant les erreurs de résolution du modèle géométrique direct d'une part et les couples actionneurs lors du retour d'effort d'autre part. Dans un deuxième temps, différentes approches ont été proposées pour résoudre les problèmes liés à la présence des singularités. La première approche a consisté à utiliser la redondance de capteurs et la redondance d'actionneurs pour palier à ces effets dans la structure existante. Dans la seconde approche, nous avons proposé une nouvelle architecture mécanique optimale qui élimine les singularités présentent dans l'espace de travail. Les résultats obtenus, avec cette nouvelle structure à travers les essais expérimentaux réalisés sur le prototype, sont conformes aux objectifs fixés. Les deux interfaces haptiques ont été utilisées pour contrôler avec succès un robot dédié à la chirurgie mini-invasive. Le comportement du système global « robot esclave interface haptique » ouvre des perspectives prometteuses aussi bien pour de futures études scientifiques que pour un transfert industriel. / A contribution for a development of haptic devices for tele-operation system is presented in this thesis. This device is dedicated for medical applications such as minimally invasive surgery tasks. In one first step, an evaluation of the existing master device is carried out. This device has a spherical parallel architecture and has been developed within the robotics team of PPRIME Institute. The evaluation of this device has shown the presence of parallel singularities located in its workspace. This singularity alters the kinematic behavior of the structure by amplifying the errors in solving the forward kinematics and amplifying the actuator torques for the haptic feedback. In a second step, different approaches have been proposed to solve the problems related to the presence of the singularities. The first approach consists in using redundancy of sensors and actuators for the existing structure in order to overcome the effects of singularities. In the second approach, we have proposed a new optimal mechanical architecture that eliminates the singularity. The results obtained with this new structure through the experimental testing of the prototype, are in accordance with the expected ones. The two haptic devices have been used to successfully control a robot dedicated to minimally invasive surgery. The behavior of the overall system "robot and haptic device" opens up promising prospects for future studies as well as for industrial transfer.

The work discussed in this thesis deals with the significance of initiator tRNA gene copy number in Escherichia coli. A summary of the relevant literature discussing the process of protein synthesis, initiator tRNA selection and gene redundancy is presented in Chapter 1. Chapter 2 describes the ‘Materials and Methods’ used in the experimental work carried out in this thesis. The next three chapters address the significance of initiator tRNA gene copy number in E. coli at three levels; at the level of the molecule (Chapter 3), at the level of the cell (Chapter 4) and at the level of the population (Chapter 5). At the end of the thesis are appended three publications, which include two papers where I have contributed to work not discussed in this thesis and one review article. A brief summary of chapters 3 to 5 is provided below: (i) Chapter 3: Can E. coli remain viable without the 3 G-C base pairs in initiator tRNA? Initiator tRNAs are distinguished from elongator tRNAs by several features key among which are the three consecutive and near universally conserved G-C base pairs found in the anticodon stem of initiator tRNAs. These bases have long been believed to be essential for the functioning of a living cell, both from in vitro and in vivo analysis. In this study, using targeted mutagenesis and an in vivo genetics based approach, we have shown that the 3 G-C base pairs can be dispensed with in E. coli, and the cell can be sustained on unconventional initiator tRNAs lacking the intact 3 G-C base pairs. Our study uncovered the importance of considering the relative amounts of molecules in a living cell, and their role in maintaining the fidelity of protein synthesis. (ii) Chapter 4: Can elongator tRNAs initiate protein synthesis? There are two types of tRNAs; initiator tRNA, of which there is one representative in the cell, and elongator tRNAs of which there are several representatives. In this study, we have uncovered initiation of protein synthesis by elongator tRNAs by depleting the initiator tRNA content in the cell. This raises the possibility that competition between initiator and elongator tRNAs at the P site of the ribosome occurs routinely in the living cell, and provides a basis for initiation at several 'start' sites in the genome that may not be currently annotated as such. We speculate that such a phenomenon could be exploited by the cell to generate phenotypic diversity without compromising genomic integrity. (iii) Chapter 5: How many initiator tRNA genes does E. coli need? E. coli has four genes that encode initiator tRNA, these are the metZWV genes that occur at 63.5 min in the genome, and the metY gene that occurs at 71.5 min in the genome. Earlier studies indicated that the absence of metY had no apparent impact on cell growth. In view of the importance of initiator tRNA gene copy number in maintaining the rate and fidelity of protein synthesis, we examined the fitness of strains carrying different numbers of initiator tRNA genes by competing them against each other in both rich and limited nutrient environments. Our results indicate a link between caloric restriction and protein synthesis mediated by the initiator tRNA gene copy number.

Dans cette thèse, nous étudions plusieurs modèles de routage efficaces en énergie. Pour chaque modèle, nous présentons une formulation en programmation linéaire mixte permettant de trouver une solution exacte. En outre, comme il s’agit de problèmes NP-Difficiles, nous proposons des heuristiques efficaces pour des réseaux de grande taille. Dans la première partie de cette thèse, nous étudions une solution de routage efficace en énergie dans laquelle nous ajoutons la possibilité d’éliminer des redondances dans les paquets transmis sur le réseau. Nous montrons premièrement que l’ajout de l’élimination des redondances permet d’améliorer l’efficacité énergétique des réseaux en éteignant plus de liens. Ensuite, nous étendons le modèle afin qu’il prenne en compte un certain niveau d’incertitudes dans le volume de trafic et le taux de redondances. La deuxième partie de cette thèse est consacrée aux problèmes qui se posent lors du déploiement de tels protocoles dans les réseaux. Plus particulièrement, nous proposons de minimiser les changements entre deux configurations réseaux consécutives lorsque plusieurs matrices de trafic sont considérées. Le routage des demandes étant alors assuré avec le protocole de routage OSPF (Open Shortest Path First). Ensuite, nous abordons le problème de la limitation du nombre de règles de routage dans les routeurs en utilisant une technologie de type SDN (Software Defined Networks). Enfin, nous présentons en annexe des travaux complémentaires réalisés au cours de cette thèse concernant le routage multicast et le contrôle de congestion TCP. / In this thesis, we study several models of energy-Aware routing. For each model, we present a linear programming formulation to find the exact solution. Moreover, since energy-Aware routing is NP-Hard problem, we also propose efficient heuristic algorithms for large scale networks. In the first part of this thesis, we deal with GreenRE - a new energy-Aware routing model with the support of redundancy elimination. We first present a deterministic model in which we show how to combine energy-Aware routing and redundancy elimination to improve energy efficiency for backbone networks. Then, we extend the model in order to take into account uncertainties in traffic volumes and redundancy rates. The second part of this thesis is devoted to the deployment issues of energy- aware routing in practice. In detail, to avoid service deterioration for end-Users, we limit changes of network configurations in multi-Period traffic matrices in Open Shortest Path First (OSPF) protocol. Next, we address the problem of limited rule space in OpenFlow switches when installing energy-Aware routing configurations. Finally, we present in the appendix other works developed during this thesis: multicast network protocol and TCP congestion control algorithm.

Submitted by Ricardo Cedraz Duque Moliterno (ricardo.moliterno@uefs.br) on 2016-02-29T22:55:34Z No. of bitstreams: 1 Disserta??o Danielle Mendes Carvalho.pdf: 1158711 bytes, checksum: b9a8e1c6f59d19bb451209ed9144c6f5 (MD5) / Made available in DSpace on 2016-02-29T22:55:34Z (GMT). No. of bitstreams: 1 Disserta??o Danielle Mendes Carvalho.pdf: 1158711 bytes, checksum: b9a8e1c6f59d19bb451209ed9144c6f5 (MD5) Previous issue date: 2013-02-04 / Funda??o de Amparo ? Pesquisa do Estado da Bahia - FAPEB / Different strategies of resource exploitation can lead to divergences in species niche realized. Niche differentiation may involve several niche dimensions, which can differ, especially in the diet, in space, and time. In this study, we measured the redundancy level or complementarity in the interaction network in a bee community associated with an agricultural system. The hypothesis is that the degree of trophic overlap on the diet axis in the community is low, with high complementarity, since there is difference in the resource supply over the vintages and, that the overlap on time axis is high with high redundancy at the time of the use of floral resources. The overlap niche on diet axis was low (NOih = 0.18) and 88% of pairs showed overlap values below 30%. Niche overlap on time axis was intermediate (NOih = 0.49), and 65% of pairs showed overlap values between 30 and 60%. The network analysis showed that the bees separate its niche on diet axis and presented additional intermediate specialization (H2 '= 0.46). And on the time axis (H2 '= 0.12) bees were considered generalists, with high temporal redundancy. When we tested whether environmental requirements determined the high redundancy at the time of use of floral resources, we verified that the temperature was not a guiding factor to structure the niche overlap in relation to the time of use. The results indicate that the low niche overlap on the diet axis and the intermediate overlap on the time axis should be reflecting the plurality of floral resources choices by different bee species, which used broad resource spectrum. / Diferentes estrat?gias de explora??o dos recursos podem levar a diverg?ncias no nicho realizado das esp?cies. Diferencia??es de nicho podem envolver diversas dimens?es do nicho, podendo diferir, sobretudo na dieta, no espa?o e no tempo. Neste estudo mensuramos o n?vel de redund?ncia ou complementaridade na rede de intera??o em uma comunidade de abelhas associada a um sistema agr?cola. Nossa hip?tese de trabalho ? que o n?vel de sobreposi??o tr?fica no eixo da dieta na comunidade seja baixo, com alta complementaridade, visto que h? diferen?a na oferta de recurso ao longo das safras e, que a sobreposi??o no eixo temporal seja alta com alta redund?ncia nos hor?rios de uso dos recursos florais. A sobreposi??o de nicho no eixo da dieta foi baixa (NOih= 0,18), e 88% dos pares apresentaram valores de sobreposi??o abaixo de 30%. A sobreposi??o de nicho no eixo temporal foi intermedi?ria (NOih= 0,49), e 65% dos pares apresentaram valores de sobreposi??o entre 30 e 60%. A an?lise de rede mostrou que as abelhas separam seu nicho no eixo da dieta e apresentaram especializa??o complementar intermedi?ria (H2?=0,46). E no eixo temporal (H2?=0,12) as abelhas foram consideradas generalistas, com alta redund?ncia temporal. Ao testarmos se requerimentos ambientais determinaram a alta redund?ncia nos hor?rios de uso dos recursos florais verificamos que a temperatura n?o foi um fator norteador para estruturar a sobreposi??o de nicho em rela??o ao hor?rio. Os nossos resultados indicam que a baixa sobreposi??o do nicho no eixo da dieta e a intermedi?ria sobreposi??o no eixo temporal devem ser reflexo da pluralidade das escolhas de recursos florais por diferentes esp?cies de abelhas, que utilizaram amplo espectro de recursos.

This master’s thesis is dealing with modification of classical discrete time wavelet transform algorithm to be able to split input image into several segments which could be independently processed. Segmented discrete time wavelet transform is introduced and implemented and the tests shows that it produces the similar values of wavelet coefficients as the classical approach. Complete independent image segment processing is allowed thanks to inverse segmented discrete time wavelet transform. It’s obvious that redundancy calculations appear when the image is processed segment by segment compared to transform the image at once. The redundancy rises with increasing decomposition depth and filter length, however it falls when the segment dimensions are increased. It happens because it is worked with smaller amount of segments which needs to extend. The extension with neighbor samples is the only source of redundancy. The thesis also contains description of two different approaches to signal border extension: firstly complete extension for all decomposition levels before the transformation itself and secondly classical method when the signal is extended in each decomposition step. Both approaches have the same results of image reconstruction, but it distinguishes in calculated wavelet coefficients. Four kinds of signal border extension are compared by using MSE and PSNR. The results of all approaches are similar, however the smooth padding of order 0 and symmetric-padding could be slightly favored against smooth padding of order 1 and zero-padding which shows higher MSE.

Tato práce se zabývá návrhem nového řídícího systému, odolného proti chybám, pro nestabilní samo-balancující dvoukolové vozidlo typu Segway. Původní systém vozidla je podroben analýze rizikovosti jeho součástí a na základě výsledků jsou navržena opatření pro zvýšení jeho bezpečnosti. Je navržena nová topologie řídícího systému obsahující dvě samostatné řídící jednotky, redundantní senzoriku a voter. Pro řídící jednotky byl vyvinut software obsahující bezpečnostní algoritmy a mechanismy přepínání kontrolních výstupů. V práci jsou také představeny dva matematické modely vozidla různé složitosti, které jsou následně využity při HIL testování nově navrženého systému. Celý návrh byl proveden s využitím nástrojů pro Rapid Control Prototyping.