• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 341
  • 133
  • 67
  • 62
  • 37
  • 22
  • 19
  • 14
  • 11
  • 8
  • 7
  • 7
  • 6
  • 5
  • 4
  • Tagged with
  • 872
  • 219
  • 99
  • 95
  • 79
  • 73
  • 68
  • 63
  • 55
  • 51
  • 49
  • 46
  • 44
  • 42
  • 41
  • 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.
261

Avaliação do custo e da efetividade da radiofrequência na dor cervical crônica com componentes autonômicos simpático e somático em um hospital público / Cost-effectiveness of radiofrequency for chronic neck pain with sympathetic and somatic components in a Public Teaching Hospital in Brazil

Zuccolotto, Veridiana Marques Rebello 29 September 2017 (has links)
Introdução: Avaliar a relação do custo e da efetividade da radiofrequência (RF) em relação ao bloqueio seqüencial na Síndrome Dolorosa Complexa Regional-tipo I (SDCR-I). Métodos: 15 pacientes com dor somática e autonômica cervical foram submetidos a bloqueio torácico simpático, combinado com bloqueio facetário cervical bilateral por 4 semanas. Esta seqüência foi repetida quando a dor atingiu VAS 4 cm, e este período foi definido como tempo de analgesia. Posteriormente, os mesmos pacientes foram submetidos a um bloqueio teste seguido de modulação de RF do gânglio simpático torácico e ablação de ramos facetários medianos cervicais. Os pacientes atuaram como seu próprio controle relacionado à analgesia, atividades rotineiras e padrão de sono. Resultados: 13 pacientes completaram o estudo. O tempo de analgesia após a sequência de 4 bloqueios foi de 4 ± 1 mês e o custo anual R$ 15.000,00. O tempo de analgesia após RF foi de 13 ± 2 meses (p <0,001) e os custos foram reduzidos em 26% no primeiro ano e 34% -38% nos anos seguintes, com extrapolação. A qualidade de vida melhorou para ambos os tratamentos (p> 0,05). Não houve efeitos adversos. Discussão: A RF resultou em analgesia de 13 meses comparada a 4 meses após os bloqueios clássicos e melhora na capacidade física e no padrão de sono. Além disso, a RF foi rentável e reduziu as taxas em 23% durante a avaliação do primeiro ano, seguida de uma redução de custos de 32% a 36% nos anos seguintes, com extrapolação. / Objectives: To evaluate cost-effectiveness of radiofrequency (RF) compared to sequential block in Complex Regional Pain Syndrome-I (CRPS). Methods: 15 patients with cervical somathic and autonomic pain were submitted to a 4- weekly sympathetic thoracic block, combined to bilateral cervical facetary block. This sequence was repeated when pain reached VAS 4-cm, and this period was defined as time of analgesia. Thereafter, same patients were submitted to a test block followed by RF modulation of thoracic sympathetic ganglion and ablation of facetary cervical median branches. Patients acted as their own control related to analgesia, routine activities, sleep pattern and costs. Results: 13 patients completed the study. The analgesia time after the 4-block sequence was 4±1 months and the annual costs USA$5000. Analgesia time after RF was 13±2 months (p<0.001) and costs were reduced by 26% in the first year and 34%-38% in the following years extrapolation. Quality of life improved for both treatments (p> 0.05). There were no adverse effects. Discussion: RF resulted in 13-month compared to 4-month analgesia after the classical 4- weekly blocks, and improved physical capacity and sleep pattern. Besides that, RF was costeffective, and reduced rates by 23% during the first-year evaluation, followed by 32%-36% cost reduction in following years, by extrapolation.
262

Entre urnas e armas: a competitividade do Poder Executivo e as Guerras Civis, 1976 - 2000 / Between Ballot Boxes and Guns: Competitiveness of Executive Branch and Civil Wars, 1976-2000

Freire, Danilo Alves Mendes 11 November 2011 (has links)
A guerra civil é a forma de violência coletiva mais importante de nossa época. Embora pesquisas recentes tenham apontado alguns elementos como determinantes das guerras civis, a influência dos fatores políticos nos conflitos internos ainda é controversa. O presente estudo analisa, por meio de regressão estatística, a relação entre a competitividade do poder executivo e a incidência de guerras civis de 1976 a 2000. Os achados indicam que tanto eleições com candidatos únicos como votações multipartidárias reduzem a incidência de guerras civis. Ademais, os resultados dão apoio às hipóteses levantadas pela literatura recente de que terreno montanhoso, grande população, sistema políticos centralizados e a existência de conflitos anteriores aumentam significativamente o risco de incidência de guerras civis / Civil War is the most important form of collective violence of our time. Although recent research has yielded some determining elements to civil war, the influence of political factors on internal conflicts remains disputed. This study analyzes, by means of statistics regression, the correlation between the competitiveness in the Executive Branch and the incidence of civil war from 1976 to 2000. The findings indicate that both single-candidate and multi-party elections reduce the incidence of civil war. Furthermore, the results lend support to the hypotheses put forward by recent literature that mountainous terrain, large population, centralized political system, and the existence of former conflicts significantly heighten the risk of incidence of civil war
263

Aplicação do método branch-and-bound na programação de tarefas em uma única máquina com data de entrega comum sob penalidades de adiantamento e atraso. / Branch-and-bound method application in a single machine earliness/tardiness scheduling problem with a common due date.

Kawamura, Márcio Seiti 07 April 2006 (has links)
O objetivo desse trabalho é o de estudar o problema de programação de tarefas num ambiente produtivo com uma única máquina com data comum de entrega. Nesse caso, as tarefas, depois de processadas uma única vez na máquina, devem ser entregues em uma data comum e sofrem penalidades de adiantamento e de atraso conforme o instante em que são completadas. Na prática, esse problema é encontrado em casos de pedidos de lotes de produtos com data de entrega comum préespecificada, embarques para exportação e material químico ou misturas que têm vida média de curta duração. Problemas desse tipo são NP-hard (Hall, Kubiak & Sethi, 1991; Hoogeven & van de Velde, 1991), sendo comumente tratados na literatura através de heurísticas e meta-heurísticas. Visto não ser de nosso conhecimento a existência na literatura de tratamento desse problema através de métodos exatos, propôs-se a utilização de um algoritmo do tipo branch-and-bound para obtenção da solução ótima do problema que minimize a soma das penalidades de adiantamento e de atraso. No desenvolvimento do algoritmo, a utilização de propriedades do problema foi importante na elaboração de limitantes inferiores e regras de dominância que melhoraram a eficiência do modelo. Os experimentos realizados avaliaram o desempenho de diferentes critérios elaborados, como escolha do nó pai, limitante inferior, ordem de execução das estratégias e ordem de construção da seqüência. Os resultados obtidos mostraram-se robustos quando comparados com o benchmark da literatura e revelaram o bom desempenho do modelo para problemas de pequeno porte, superando o desempenho de programas de otimização comerciais. / The objective of this work is to study the single-machine scheduling problem with a common due date. In this case, jobs, after be processed only once in the machine, must be delivered in a common due date and they are penalized of earliness or tardiness according to their completion time. This problem is found in cases of batch production with prespecified common due date, exportation shipping and chemical material that has short half-life period. This kind of problem is NP-hard (Hall, Kubiak & Sethi, 1991; Hoogeven & van de Velde, 1991) and it has been treated in the literature by heuristics and meta-heuristics. Not having knowledge about previous treatment by exact methods in the literature, it was proposed the implementation of a branch-and-bound algorithm to obtain the optimal solution that minimizes the total weighted earliness and tardiness penalties. In the development of the algorithm, the utilization of problem properties was important to the elaboration of lower bounds and pruning rules that have enhanced the efficiency of the model. The realized tests have evaluated the performance of different criteria, like the choice of father node, lower bound, strategy execution order and sequence construction order. The obtained results have demonstrated robustness comparing to benchmark and they have revealed the good working of the model for small problems, overcoming optimization software performance.
264

Regionalization of Scar Patterns on the Florida Manatee (Trichechus manatus latirostris) Observed at Harbor Branch Oceanographic Institute, Florida.

Unknown Date (has links)
The Florida manatee (Trichechus manatus latirostris) is native to Florida and the Indian River Lagoon. Harbor Branch Oceanographic Institute (HBOI) is located in Indian River Lagoon and is frequently visited by manatees. The Manatee Project was created in 2009 to document and photograph the manatees visiting HBOI. Analyzing photographs of 146 manatee that visited HBOI showed that a majority of the injuries sustained were caused by boats. 97% of the manatee had at least one propeller injury and 31% of the manatee had at least one skeg injury. Other non-boat related injuries seen in the images included cold stress and entanglement injuries. This study looked at the prevalence of scar by anatomical region, the cause of injury, and compared injury locations between male and female manatees. / Includes bibliography. / Thesis (M.S.)--Florida Atlantic University, 2018. / FAU Electronic Theses and Dissertations Collection
265

A gestão financeira e sua implicação no desempenho do negócio dos empreendedores no ramo da beleza da cidade de Sant’Ana do Livramento-RS / The financial management and its implication in the performance of the entrepreneurs business in the beauty sector of the city of Sant'Ana do Livramento-RS

Barroso, Roseclair da Rocha Lacerda 16 March 2018 (has links)
Submitted by Karen Britto (karenbritto@unipampa.edu.br) on 2018-05-23T15:16:17Z No. of bitstreams: 1 Roseclair da Rocha Lacerda Barroso - 2018.pdf: 1345467 bytes, checksum: 3a3ae9c9ef6abab66cfd96aaa4c37b26 (MD5) / Made available in DSpace on 2018-05-23T15:16:17Z (GMT). No. of bitstreams: 1 Roseclair da Rocha Lacerda Barroso - 2018.pdf: 1345467 bytes, checksum: 3a3ae9c9ef6abab66cfd96aaa4c37b26 (MD5) Previous issue date: 2018-03-16 / Com o foco voltado para a gestão dos pequenos negócios e na importância do empreendedor para a economia do país, é que esta dissertação teve o objetivo de avaliar a gestão financeira e sua implicação no desempenho do negócio dos empreendedores no ramo da beleza na cidade de Sant’ana do Livramento/RS. No intuito de atingir os objetivos propostos, foi desenvolvida uma pesquisa de caráter descritivo com abordagem qualitativa, conduzida pelo método da Pesquisa-ação, constituída por três etapas: entrevista inicial, experiência de controle financeiro e entrevista final, realizadas entre agosto e dezembro de 2017 com quatro empreendedoras do ramo da beleza, três atuantes na atividade de cabeleireira e uma nas atividades de manicure, pedicure e depiladora. O método de pesquisa-ação mostrou-se propício à realização da pesquisa, pois firmou a intenção de auxiliar as empreendedoras na detecção de problemas relacionados à gestão financeira e também a outros assuntos inerentes à administração dos negócios. Ao caracterizar as empreendedoras e seus negócios, foram identificadas microempreendedoras proprietárias de salão de beleza, que já possuíam familiaridade com o ramo e iniciaram seus empreendimentos com capital próprio, tendo-os como única fonte de renda. Como resultados, foi constatado inicialmente, que três das quatro empreendedoras não conheciam a gestão financeira dos seus negócios, desconhecendo desde os valores de entradas e saídas de dinheiro em caixa até os resultados financeiros. Com base na entrevista inicial foi proposta e realizada uma experiência de controle financeiro durante o período de três meses, dentro dos salões de beleza das empreendedoras que possibilitou, segundo os relatos das participantes, uma melhor organização e maior visão dos resultados financeiros, o que evidencia a adequação das ações à realidade dos seus negócios, no momento da realização da pesquisa. Também foi possível averiguar ao final da experiência, por meio da observação participante, que as empreendedoras demonstraram maior consciência quanto à importância de se conhecer os resultados financeiros para a tomada de decisões. Quanto ao desempenho, foi evidenciado que, sem uma sistemática para se efetuar registros e sem métodos de controle, não se conhece os números da empresa e, por consequência o desempenho financeiro do negócio. Sendo assim, pôde-se concluir que a gestão financeira implica no desempenho dos negócios de forma significativa, no caso das empreendedoras participantes. Porém, a falta de conhecimento sobre finanças não foi fator determinante para o estabelecimento dos seus negócios no mercado. Os resultados obtidos nesta pesquisa podem contribuir para a extrapolação desta experiência em outros pequenos negócios. / With the focus on the management of small businesses and the importance of the entrepreneur to the country's economy, this dissertation aimed to evaluate financial management and its implication in the business performance of entrepreneurs in the beauty sector in the city of Sant 'ana do Livramento / RS. In order to reach the proposed objectives, a descriptive research was developed with a qualitative approach, conducted by the Action Research method, consisting of three stages: initial interview, experience of financial control and final interview, carried out between August and December 2017 with four beauty entrepreneurs, three in the hairdressing activity and one in the activities of manicure, pedicure and epilator. The Action Research method proved to be conducive to carrying out the research, since it established the intention of assisting entrepreneurs in the detection of problems related to financial management and also to other matters inherent to business administration. In characterizing entrepreneurs and their businesses, microentrepreneurs who owned beauty salons were identified, who already had familiarity with the industry and started their entrepreneurship with their own capital, having them as sole source of income. As a result, it was found initially that three of the four entrepreneurs did not know the financial management of their businesses, ignoring from the values of inflows and outflows of cash to financial results. Based on the initial interview, a financial control experience was proposed and carried out during the three-month period, within the beauty salons of the entrepreneurs, which enabled, according to the participants' reports, a better organization and greater vision of the financial results, which evidences the adequacy of actions to the reality of their business, at the time of the research. It was also possible to verify at the end of the experience, through participant observation, that the entrepreneurs showed greater awareness of the importance of knowing the financial results for decision making. Regarding performance, it was evidenced that, without a systematic to carry out records and without control methods, the numbers of the company and, consequently, the financial performance of the business are not known. Thus, it can be concluded that financial management implies a significant performance of the business, in the case of participating entrepreneurs. However, the lack of knowledge about finance was not a determining factor for the establishment of its business in the market. The results obtained in this research can contribute to the extrapolation of this experience in other small businesses.
266

Estimação de estado para redes de distribuição de energia elétrica avançadas / State estimation for advanced electric power distribution networks

Ismail Neto, Mohamad Salem 26 May 2017 (has links)
O estimador de estado (EE) foi originalmente desenvolvido para lidar com sistemas de transmissão (ST). Desta maneira, sua aplicação em sistemas de distribuição (SD) exige que os métodos clássicos sejam repensados. Este trabalho teve como objetivo avaliar os métodos disponíveis na literatura para a estimação de estado em SD e explorar a fundo um método escolhido. A abordagem escolhida foi a que utiliza as correntes complexas nos ramos como variáveis de estado. Tal abordagem visa tirar proveito das características específicas dos SD, e apresenta um desempenho computacional superior se comparado a soluções que utilizam as tensões como variáveis de estado. Entretanto, tal método ainda se encontra pouco explorado na literatura, de modo que diversas limitações foram identificadas. A principal limitação encontrada foi no que tange à presença de transformadores no sistema, que foi negligenciada nos trabalhos encontrados na literatura. De modo a aumentar a confiabilidade da estimação, além das correntes em todos os ramos, foi considerada a possibilidade de se incluir no vetor de estado variáveis referentes à tensão na subestação (SE). Na presença de transformadores, tal expansão no vetor de estados gerou um acoplamento do modelo matemático. Neste trabalho, foram propostas soluções para lidar com diferentes conexões de transformadores utilizando o estimador com base nas correntes mantendo uma de suas principais características, que é o desacoplamento entre fases. Na ocasião da estimação da tensão na SE, um método de combinação ótima de variáveis de estado foi proposto de modo a contornar o acoplamento gerado no modelo matemático. Sistemas teste do IEEE foram utilizados para avaliar o desempenho do algoritmo implementado. Os resultados obtidos mostraram que a metodologia escolhida foi capaz de estimar o estado do SD com precisão adequada na presença de diferentes tipos de medidas e topologias. Os modelos propostos neste trabalho se mostraram eficazes na incorporação de transformadores com diferentes conexões na formulação do EE. Por fim, o método proposto para lidar com a estimação da tensão na SE de maneira desacoplada na presença de transformadores foi validado e se mostrou mais eficiente computacionalmente quando comparado ao seu análogo acoplado. / The state estimator (SE) was originally developed to deal with transmission networks (TN). Hence, its application in distribution networks (DN) requires the classical methods to be modified. The main aim of this work was to evaluate the methods available in the literature for state estimation in DN and to explore in depth a chosen method. The chosen approach was the one that uses the complex branch currents as state variables. This approach aims to take advantage of the specific characteristics of DN, and presents a superior computational performance when compared to solutions that use voltages as state variables. However, this method is still little explored in the literature, so that several limitations have been identified. The main limitation was in terms of the presence of transformers in the system, which was neglected in the papers found in the literature. In order to increase the reliability of the estimation, in addition to the branch currents, it was considered the posibility of including the substation voltage variables in the state vector. In the presence of transformers, such expansion in the state vector generated a coupled mathematical model. In this work, solutions were proposed to deal with different transformer connections using the branch current SE while keeping one of its main characteristics, that is the decoupling between phases. When the substation voltage is being estimated, an optimal combination of state variables was proposed in order to circumvent the coupling in the mathematical model. IEEE test systems were used to evaluate the performance of the implemented algorithm. The results showed that the chosen methodology was able to estimate the DN state with adequate accuracy in the presence of different types of measurements and topologies. The models proposed in this work proved to be effective in the incorporation of transformers with different connections in the SE formulation. Finally, the proposed method to decouple the branch current SE when dealing with substation voltage estimation in the presence of transformers was validated and proved to be computationally more efficient when compared to its coupled version.
267

The design of PABX with LAN architecture.

January 1992 (has links)
Ko Wing Hoi. / Duplicate numbering of leave 67. / Thesis (M.Sc.)--Chinese University of Hong Kong, 1992. / Includes bibliographical references (leaves 71-72). / Chapter 1. --- INTRODUCTION --- p.1 / Chapter 2. --- COMPARISONS OF LAN AND PABX --- p.3 / Chapter 2.1 --- Typical LAN system --- p.3 / Chapter 2.1.1 --- Characteristics of a LAN [1] --- p.3 / Chapter 2.1.2 --- Transmission medium of LAN --- p.5 / Chapter 2.1.3 --- LAN access control methods --- p.6 / Chapter 2.1.4 --- Interfacing to the LAN --- p.8 / Chapter 2.1.5 --- LAN topology --- p.8 / Chapter 2.1.6 --- Switching techniques --- p.9 / Chapter 2 .2 --- Applications of LAN --- p.11 / Chapter 2.2.1 --- Small filestore LAN's --- p.12 / Chapter 2.2.2 --- Wiring replacement LAN's --- p.12 / Chapter 2.2.3 --- Personal computer networks --- p.13 / Chapter 2.2.4 --- General purpose LAN's --- p.13 / Chapter 2 .3 --- Typical PABX system --- p.14 / Chapter 2.3.1 --- PABX topology --- p.15 / Chapter 2.3.2 --- Circuit switching --- p.15 / Chapter 2.3.3 --- Telephony signalling --- p.16 / Chapter 2.3.3.1 --- Pulsing --- p.16 / Chapter 2.3.3.2 --- Subscriber loop signaling [2] --- p.17 / Chapter 2.3.4 --- ISDN (Integrated Services Digital Network) --- p.19 / Chapter 2.4 --- Applications of PABX --- p.21 / Chapter 2.5 --- Comparisons of LAN and PABX --- p.22 / Chapter 3. --- INTEGRATION OF PABX WITH LAN --- p.25 / Chapter 3.1 --- Advantages of integration of PABX with LAN --- p.25 / Chapter 3.1.1. --- LAN-PABX Gateway --- p.28 / Chapter 3.1.2. --- Problems in interconnecting PABX and LAN [6] --- p.29 / Chapter 3.1.3. --- ISDN-PABX [7] --- p.30 / Chapter 3.2 --- Architecture of Integrated LAN and PABX --- p.31 / Chapter 3.3 --- Typical applications --- p.32 / Chapter 4. --- CALL PROCESSING --- p.35 / Chapter 4.1 --- Finite State Diagrams for voice calls --- p.37 / Chapter 4.2 --- SDL representations of voice calls --- p.39 / Chapter 4.3 --- Software implementations of SDL diagrams --- p.40 / Chapter 4.3.1 --- PABX operating system --- p.40 / Chapter 4.3.2 --- Trunk operating system --- p.43 / Chapter 4.3.3 --- Message format --- p.43 / Chapter 4.4 --- Pseudo codes for PABX --- p.45 / Chapter 4.4 --- Pseudo codes for trunks --- p.52 / Chapter 5. --- HARDWARE IMPLEMENTATION --- p.57 / Chapter 5.1 --- TRUNK INTERFACE --- p.58 / Chapter 5.1.1 --- PABX to CO call --- p.58 / Chapter 5.1.2 --- CO to PABX call --- p.59 / Chapter 5.2 --- Subscriber Interface Circuit --- p.59 / Chapter 5.4 --- PSTN Trunk Interf ace --- p.60 / Chapter 6. --- CONCLUSIONS --- p.62 / Acknowledgements --- p.64 / APPENDIX A --- p.65 / CCITT SPECIFICATION AND DESCRIPTION LANGUAGE [15] --- p.65 / APPENDIX B --- p.68 / "SIGNALLING FOR SWITCHING SYSTEMS IN HK [16],[17]" --- p.68 / Chapter B. 1 --- Tone plan --- p.68 / Chapter B. 2 --- Tone levels --- p.68 / Chapter B. 3 --- Ringing frequency and voltage --- p.68 / Chapter B. 4 --- Dial pulse --- p.68 / Chapter B. 5 --- DTMF (Dual-tone multi-frequency) --- p.69 / Chapter B. 6 --- PCM coding --- p.69 / REFERENCES --- p.71
268

The antecedents and consequences of student perceptions of university image and student-university identification in transnational higher education

Wilkins, Stephen Joseph Karl January 2013 (has links)
This research aims to identify the process by which students form images of universities, the extent to which students’ favourable evaluations of image attractiveness lead to student-university identification, and the extent to which perceived image attractiveness and student-university identification determine planned behaviour, i.e., supportive intentions, including student choice of institution. Full-service international branch campuses offering complete degree programmes are a fairly new phenomenon on the higher education landscape and potential students have limited knowledge about them and the institutions that own them. It is interesting therefore to discover whether these students do in fact hold images of international branch campuses. The research was conducted in the United Arab Emirates (UAE), the country that hosts more international branch campuses than any other worldwide. The study adopted a deductive, quantitative method, which involved a survey questionnaire completed by potential university students (year 12 and 13 high school students). This research stands out from earlier work on organisational identification, as earlier studies focused on existing consumers or employees while this study considers potential consumers (students). The research included a pilot study that involved individual interviews with members of the target population, which ensured research design validity. Data were analysed using a variety of techniques including exploratory factor analysis, multiple regression and structural equation modelling. The findings of this study provide support for the proposition that individuals can identify with universities in the absence of formal membership – with no or minimal previous interaction between the individual and the university – and that student-university identification can lead to supportive intentions among prospective students. These findings suggest that institutions would benefit from articulating and communicating their identities clearly, coherently and in a persuasive manner, and emphasising those aspects of the university’s identity that prospective students will perceive as prestigious, distinctive and similar to their own identities.
269

Analyse réaliste d'algorithmes standards / Realistic analysis of standard algorithms

Auger, Nicolas 20 December 2018 (has links)
À l'origine de cette thèse, nous nous sommes intéressés à l'algorithme de tri TimSort qui est apparu en 2002, alors que la littérature sur le problème du tri était déjà bien dense. Bien qu'il soit utilisé dans de nombreux langages de programmation, les performances de cet algorithme n'avaient jamais été formellement analysées avant nos travaux. L'étude fine de TimSort nous a conduits à enrichir nos modèles théoriques, en y incorporant des caractéristiques modernes de l'architecture des ordinateurs. Nous avons, en particulier, étudié le mécanisme de prédiction de branchement. Grâce à cette analyse théorique, nous avons pu proposer des modifications de certains algorithmes élémentaires (comme l'exponentiation rapide ou la dichotomie) qui utilisent ce principe à leur avantage, améliorant significativement leurs performances lorsqu'ils sont exécutés sur des machines récentes. Enfin, même s'il est courant dans le cadre de l'analyse en moyenne de considérer que les entrées sont uniformément distribuées, cela ne semble pas toujours refléter les distributions auxquelles nous sommes confrontés dans la réalité. Ainsi, une des raisons du choix d'implanter TimSort dans des bibliothèques standard de Java et Python est probablement sa capacité à s'adapter à des entrées partiellement triées. Nous proposons, pour conclure cette thèse, un modèle mathématique de distribution non-uniforme sur les permutations qui favorise l'apparition d'entrées partiellement triées, et nous en donnons une analyse probabiliste détaillée / At first, we were interested in TimSort, a sorting algorithm which was designed in 2002, at a time where it was hard to imagine new results on sorting. Although it is used in many programming languages, the efficiency of this algorithm has not been studied formally before our work. The fine-grain study of TimSort leads us to take into account, in our theoretical models, some modern features of computer architecture. In particular, we propose a study of the mechanisms of branch prediction. This theoretical analysis allows us to design variants of some elementary algorithms (like binary search or exponentiation by squaring) that rely on this feature to achieve better performance on recent computers. Even if uniform distributions are usually considered for the average case analysis of algorithms, it may not be the best framework for studying sorting algorithms. The choice of using TimSort in many programming languages as Java and Python is probably driven by its efficiency on almost-sorted input. To conclude this dissertation, we propose a mathematical model of non-uniform distribution on permutations, for which permutations that are almost sorted are more likely, and provide a detailed probabilistic analysis
270

A branch-and-price algorith, for a compressor scheduling problem

Friske, Marcelo Wuttig January 2016 (has links)
O presente trabalho realiza o estudo e aplicação de um algoritmo de branch-and-price para a resolução de um problema de escalonamento de compressores. O problema é ligado à produção petrolífera, o qual consiste em definir um conjunto de compressores a serem ativados para fornecer gas de elevação a um conjunto de poços, atendendo toda demanda e minimizando os custos envolvidos. O problema é caracterizado por uma função objetivo não-convexa que é linearizada por partes de forma a ser formulada como um problema de programação inteira mista. A abordagem de geração de colunas é baseada na decomposição de Dantzig-Wolfe e apresenta melhores limitantes inferiores em relação à relaxação linear da formulação compacta. O branch-and-price é comparado ao solver CPLEX, sendo capaz de encontrar a solução ótima em menor tempo para um conjunto de instâncias, bem como melhores soluções factíveis para instâncias maiores em um período de tempo limitado. / This work presents the study and application of a branch-and-price algorithm for solving a compressor scheduling problem. The problem is related to oil production and consists of defining a set of compressors to be activated, supplying the gas-lift demand of a set of wells and minimizing the associated costs. The problem has a non-convex objective function, to which a piecewise-linear formulation has been proposed. This dissertation proposes a column generation approach based on the Dantzig-Wolfe decomposition, which achieves tighter lower bounds than the straightforward linear relaxation of the piecewise-linear formulation. The column generation was embedded in a branch-and-price algorithm and further compared with CPLEX, obtaining optimal solutions in lesser time for a set of instances. Further, the branch-and-price algorithm can find better feasible solutions for large instances under a limited processing time.

Page generated in 0.0805 seconds