Plus de la moitié des pertes de bois dans les forêts européennes sont dues aux tempêtes. Une connaissance des mécanismes impliqués dans la stabilité mécanique des arbres est alors capitale. L’ancrage de l’arbre dans le sol constitue l’une des composantes principales du maintien mécanique de l’arbre. Il est principalement déterminé par l’architecture du système racinaire et son interaction mécanique avec le sol. Au cours de son développement, l’arbre modifie ses dimensions et se complexifie. Plus particulièrement, le système racinaire semble s’acclimater (ex : croissance en diamètre plus importante) aux déformations engendrées par le vent. L’ensemble de ces modifications conduit à une évolution des mécanismes à l’origine de l’ancrage au cours du développement de l’arbre. L’étude expérimentale de cette fonction est compliquée car les racines sont difficilement mesurables en continu dans le sol. Nous avons alors mis au point une approche numérique pour décrire la croissance du système racinaire et la distribution des déformations dues au vent. Une grande base de données structurée en chronoséquence de systèmes racinaires numérisés (Pinus pinaster) a été mobilisée. Comme l’étude de la structure et des fonctions des racines est plus efficiente quand la différentiation entre racines est prise en compte, nous avons d’abord formalisé les types racinaires du système racinaire du pin maritime à partir d’une technique de classification (« k-means clustering ») réalisée avec quatre variables. La classification des racines latérales du pin maritime nous a permis d’identifier 5 types racinaires au cours du développement du pin maritime. Ce regroupement explique 70% de la variabilité de notre base de données. Chaque système racinaire est caractérisé par trois grosses racines horizontales émises par la souche. Les racines montrent une forte différentiation pour leur tropisme, avec une direction de croissance soit horizontale soit verticale. La structure de la partie centrale du système racinaire est pratiquement complète dès l'âge de 4 ans. Sur la base des types racinaires identifiés, nous avons calibré un modèle architectural (RootTyp ; Pagès et al. 2004) pour le pin maritime. Treize paramètres pour chaque type racinaire ont été estimés par l’intermédiaire de la base de données, d’informations issues de la littérature et d’une procédure d'optimisation. Une modélisation réaliste du système racinaire jusqu'à 50 ans n’a pu être obtenue qu'en implémentant au modèle RootTyp de nouveaux processus biologiques : la diminution de la ramification avec la croissance de la racine et la diminution de la vigueur des racines avec l'ordre de ramification. Malgré ces améliorations, les systèmes racinaires de la base de données présentent des diamètres plus importants à proximité de la souche par rapport aux systèmes racinaires simulés. Ce biais systématique est principalement attribué à l’acclimatation des racines au vent dominant. Les altérations de croissance dues aux contraintes pédologiques ont également été implémentées grâce à l’amélioration du module de sol du modèle architectural.Enfin, pour comprendre les mécanismes à l’origine de l’acclimatation des racines nous avons combiné plusieurs modèles pour prédire la distribution spatiale des déformations dans des maquettes simplifiées de systèmes racinaires à 4, 6 et 13 ans, pour trois régimes de vent spécifiques à la région étudiée. D’après les simulations, les déformations des racines sous l'effet du vent diminuent avec l’âge, en raison de l’augmentation de la rigidité des racines. Cela suggère une plus forte réponse thigmomorphogénétique aux stades jeunes. Les modifications structurelles et anatomiques du système racinaire par acclimatation au vent s’expliquent principalement par les distributions des déformations et des contraintes dans les racines. / Storms cause more than 50% of the timber loss in European forests. However, forest tree anchorage mechanisms throughout their lifespan are not fully understood, especially the strong acclimation of root systems to common winds. This lack of knowledge is mainly due to technical difficulties: neither the root structure nor the mechanical contribution of the roots could be characterized continually. Thus we set up a numerical approach to model the development of the root system and to describe the strains resulting from common winds. This generic approach has been developed using Pinus pinaster grown in sandy soils as model species.Seven datasets of excavated root systems from 0 to 50 years were employed. The assessment of root structure and functions is more powerful if the differentiation of root system in several root types is considered. We first proposed an automatic classification of roots with the k-means clustering algorithm. Four root traits were chosen as classifiers, including three geometric architectural traits, which can be precisely assessed whatever the tree/root age. Clustering yielded similar five groups of laterals roots at all ages, explaining 70% of the variability. The three largest lateral roots per tree were all horizontal roots branching from stump and the other lateral roots show a large differentiation for tropism: nearly all the roots were horizontal or vertical roots. The framework of the central part of the root system can be almost completed in 4-year-old trees (3.5 cm collar diameter). We then calibrated the existing RootTyp (Pagès et al. 2004) architectural model for P. pinaster for each of the root types defined by the cluster analysis. We used the database combined with a literature review and an optimization method to get accurate values for 13 parameters by root types. We devoted effort to validate our model calibration. In order to model architecture of the root system, damping properties had to be implemented to yield realistic outputs up to the mature stage. Branching varied as a function of distance from the root base, and growth capacity decreased with branching order. Nevertheless, the root diameters of simulated root systems were generally underestimated. This was certainly due to root growth plasticity to the prevailing wind, an acclimation facet not taken into account at this calibration step. Growth alterations due to a cemented horizon were reproduced using the new calibrated soil module. Then, the wind acclimation of roots was numerically investigated by examining the root mechanical stimuli due to wind. A chain of biomechanical models was used to predict the spatial distribution of stress and strain in simplified root systems at 4, 6 and 13-year-old as a result of three levels of usual winds. According to simulations, the strain amplitude decreased with tree growth due to the increasing root system stiffness. This suggests larger thigmomorphogenetic responses at young stages. The modifications of the structural and wood root properties related to wind acclimation were largely explained by the stress and strain distribution in the root system.

Esta tese visa demonstrar a aplicação do padrão ODP (Open Distributed System) no desenvolvimento de sistemas distribuídos e abertos aplicados em automação. Para atingir este objetivo define-se uma estratégia cujos elementos básicos são: a metodologia ODP e o experimento ODP. A metodologia suporta o padrão ODP, numa estrutura sistemática e o experimento, aplica-se num projeto, para observar o comportamento do processo de aplicabilidade. A metodologia ODP está composta dos conceitos, procedimentos e modelos, propostos pelo padrão ODP; das técnicas e conceitos da orientação a objetos e da estrutura de um modelo de sistema aberto de automação (SAA), constituindo uma estrutura coerente e gerenciável, de aplicação prática nos projetos de sistemas abertos em automação. No experimento ODP, a metodologia é aplicada no desenvolvimento de um sistema de automoção, cujo escopo são os projetos dos grandes empreendimentos, que utilizam em seu processo de desenvolvimento a engenharia simultânea. O resultado desta automação será caracterizada pela obtenção de um sistema distribuído, aberto e configurado como ambiente cooperativo (Groupware) de projeto. Finalmente, os resultados do experimento serão comparados com outros tipos de metodologias que não utilizam o ODP, para poder concluir sobre a utilização do padrão ODP nos projetos de sistemas de automação. / This thesis purposes to demonstrate the applicability of the open distributed processing standard, in the development of the open distributed system applied in the automation. To achieve this objective a strategy is defined, whose basic elements are the ODP methodology and the ODP experiment. The methodology supports the ODP standard in a systematic structure and the experiment applies it in one project, to observe the behavior of the applicability process. The ODP methodology is composed by concepts, procedures and models, proposed by the ODP standard; by techniques and concepts of the objects oriented, and the structure of an open automation system model (OAS), establishing an manageable and coherent structure. In the ODP experiment, this methodology is applied in the development of an automation system which scope is the large enterprise projects that use in his development process the concurrent engineering. The result of this automation will be characterized by the Open Distributed System and will be configured as a project cooperative environment (Groupware). Finally, the experiment results will be compared with other methodologies that don\'t use ODP, to obtain conclusions about the use of the ODP standard in the automation system projects.

Esta tese visa demonstrar a aplicação do padrão ODP (Open Distributed System) no desenvolvimento de sistemas distribuídos e abertos aplicados em automação. Para atingir este objetivo define-se uma estratégia cujos elementos básicos são: a metodologia ODP e o experimento ODP. A metodologia suporta o padrão ODP, numa estrutura sistemática e o experimento, aplica-se num projeto, para observar o comportamento do processo de aplicabilidade. A metodologia ODP está composta dos conceitos, procedimentos e modelos, propostos pelo padrão ODP; das técnicas e conceitos da orientação a objetos e da estrutura de um modelo de sistema aberto de automação (SAA), constituindo uma estrutura coerente e gerenciável, de aplicação prática nos projetos de sistemas abertos em automação. No experimento ODP, a metodologia é aplicada no desenvolvimento de um sistema de automoção, cujo escopo são os projetos dos grandes empreendimentos, que utilizam em seu processo de desenvolvimento a engenharia simultânea. O resultado desta automação será caracterizada pela obtenção de um sistema distribuído, aberto e configurado como ambiente cooperativo (Groupware) de projeto. Finalmente, os resultados do experimento serão comparados com outros tipos de metodologias que não utilizam o ODP, para poder concluir sobre a utilização do padrão ODP nos projetos de sistemas de automação. / This thesis purposes to demonstrate the applicability of the open distributed processing standard, in the development of the open distributed system applied in the automation. To achieve this objective a strategy is defined, whose basic elements are the ODP methodology and the ODP experiment. The methodology supports the ODP standard in a systematic structure and the experiment applies it in one project, to observe the behavior of the applicability process. The ODP methodology is composed by concepts, procedures and models, proposed by the ODP standard; by techniques and concepts of the objects oriented, and the structure of an open automation system model (OAS), establishing an manageable and coherent structure. In the ODP experiment, this methodology is applied in the development of an automation system which scope is the large enterprise projects that use in his development process the concurrent engineering. The result of this automation will be characterized by the Open Distributed System and will be configured as a project cooperative environment (Groupware). Finally, the experiment results will be compared with other methodologies that don\'t use ODP, to obtain conclusions about the use of the ODP standard in the automation system projects.

Made available in DSpace on 2016-04-29T14:23:28Z (GMT). No. of bitstreams: 1 Jose Roberto Madureira Junior.pdf: 2597024 bytes, checksum: f6b607352cabb213c3e453dae6b3b008 (MD5) Previous issue date: 2014-03-14 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The following research focuses on the widespread use of Cloud Computing as a result of a major market boost due to the latest trends and needs. It has progressed as a result of advanced technology providing businesses a high standard of service as well as being of great benefit to its users. This final work is aimed at demonstrating users and potencial ones the benefits as well as the possibilities and challenges involved in the use of Cloud Computing and therefore allow them to identify, understand and manage such a service. As a result users shall be able to hire services involving a low risk factor in addition to implementing an appropriate workload. Due to the lack of information regarding Cloud Computing services, we believe that this research can provide users a valuable contribution in clarifying and highlighting both the potentiality and benefits involved as well as analysing the reasons they have not been able to benefit from them so far. We would like to point out that when we searched for literature regarding specifically the service users no significant data was collected. As a result we strongly believe that if such material is published it will represent a significant contribution towards both the users and the academic community. At first we carried out a state of the art study regarding Cloud Computing, identifying the potential benefits involved in providing such a service as well as the impact such a model has had on the information technology scenario. Then we implemented the case study metodology in which three real cases with the model in use were closely followed. Reference is also made to prestigious companies with a solid background within their market segment and the respective services they provide. This will allow the users to obtain the necessary backup in order to be able to previously analyse the ideal sevice which will suit their needs. We concluded that companies seem to be cautious about implementing such a structure due to their concern about mainly operational and safety measures. Cloud Computing´s progress depends on such an acceptance. Therefore this text focuses on identifying and analysing the challenges and benefits involved in implementing such a model. Research has been done with the specific aim of offering a wide range of businesses an overview of Cloud Computing as well as enabling users to overcome any difficulties / Esta pesquisa trata da utilização de Cloud Computing, mostrando sua disseminação impulsionada pelas tendências e necessidades do mercado. Sua evolução é resultado dos avanços tecnológicos que possibilitam e garantem a disseminação e solidez dos serviços prestados, oferecendo grandes benefícios aos usuários. O objetivo desta dissertação é demonstrar as potencialidades, possibilidades e desafios da utilização de serviços de Cloud Computing para que usuários e potenciais usuários possam identificá-los, compreendê-los e prepararem-se para gerenciá-los, de modo tal a poder, nestes ambientes, contratar serviços com grau aceitável de risco e implementar a carga de trabalho correta. A falta de informações sobre os serviços de Cloud Computing. Nos leva a acreditar que esta pesquisa pode oferecer uma contribuição de alto valor para o esclarecimento dos potenciais e benefícios a serem alcançados com esta tecnologia, bem como analisar os desafios que impedem tais benefícios de serem alcançados pelos seus usuários. No que diz respeito especificamente aos usuários destes serviços, na revisão de literatura realizada nada de valor significativo foi encontrado. Desta forma acreditamos que produção e disponibilização de material esclarecedor e útil relacionado ao foco no usuário e demais assuntos relacionados ao tema proposto à comunidade acadêmica seja uma importante contribuição. Para tal, primeiramente, realizamos um estudo do estado da arte de Cloud Computing, identificando os potenciais benefícios obtidos na adoção deste serviço, bem como, a exposição de como o modelo tem transformado o cenário da tecnologia da informação. Em seguida, adotamos a metodologia de estudo de casos onde observamos três casos reais da utilização de serviços do modelo. São destacadas também empresas de renome e seus respectivos serviços oferecidos, que são relevantes em função de suas história e solidez no mercado, possibilitando ao usuário uma análise prévia para a aquisição do serviço ideal ao seu ramo de atividade. Concluímos que a adoção deste tipo de estrutura é tratada com cautela pelas empresas, por conta de preocupações quanto à interoperabilidade, à segurança e os acordos de nível de serviço, entre outros. Tais questões estão ditando o ritmo em que a Cloud Computing vem sendo adotada, bem como sua evolução. Desta forma, neste texto, identificamos e analisamos os desafios que impedem e dificultam a adoção dos benefícios deste modelo. Por fim, nossa intenção foi oferecer uma visão geral dos serviços de Cloud Computing que estão sendo utilizados e que ajudam aos usuários a vencer desafios em diferentes áreas

Access control is a significant issue in any secure computer system. Authorization models provide a formalism and framework for specifying and evaluating access control policies that determine how access is granted and delegated among particular users. The aim of this dissertation is to investigate flexible decentralized authorization model supporting authorization delegation, both positive and negative authorization, and conflict resolution. A graph based authorization framework is proposed which can support authorization delegations and both positive and negative authorizations. In particular, it is shown that the existing conflict resolution methods are limited when applied to decentralized authorization models and cyclic authorizations can even lead to undesirable situations. A new conflict resolution policy is then proposed, which can support well controlled delegation by giving predecessors higher priorities along the delegation path. The thesis provides a formal description of the proposed model and detailed descriptions of algorithms to implement it. The model is represented using labelled digraphs, which provide a formal basis for proving the semantic correctness of the model. A weighted graph based model is presented which allows grantors to further express degrees of certainties about their granting of authorizations. The work is further extended to consider more complex domains where subjects, objects and access rights are hierarchically structured and authorization inheritance along the hierarchies taken into account. A precise semantics is given which is based on stable model semantics, and, several important properties of delegatable authorization programs investigated. The framework provides users a reasonable method to express complex security policy. To address the many situations in which users may need to be granted or delegated authorizations for a limited period of time, a temporal decentralized authorization model is proposed in which temporal authorization delegations and negations are allowable. Proper semantic properties are further investigated. Finally, as an application, the thesis shows how the proposed authorization model can be used in a e-consent system on health data. A system architecture for e-consent is presented and different types of e-consent models discussed. The proposed model is shown to provide users a good framework for representing and evaluating these models. / Doctor of Philosphy (PhD)

As IT industries develop, upgrade and diversify, heterogeneous operating environments running a mix of new and legacy systems become increasingly important. Such environments are currently inadequate due to lack of compatibility with each other. This thesis investigates how agents can be utilised to facilitate such heterogeneous environments, aiding enterprise systems in building bridges between incompatible software and hardware systems. An autonomous agent has independent agency and decision-making astuteness. When placed in heterogeneous environments to interact with other such agents, the consequences of its action and its preferred choice of actions are greatly influenced by actions of other agents interacting in heterogeneous environments. The main objectives of this thesis include examining the roles of agents in heterogeneous operating environments, development of a novel multi agent base architecture and an associated framework for single and heterogeneous environment. The research work also studies the plausible application to test the developed proof of concept by developing application and using the framework that utilises Windows services in a totally incompatible Solaris based Sun Ray ultra thin client environment. The work includes a novel method of modeling agent based communication architecture suitable for correspondence between two inherently different operating systems - Solaris and Microsoft Windows. The circumstances in which coordination or coordination failure occurs between these systems are investigated. The proposed method of agent based communication that can potentially overcome the barriers formed by two completely different software and hardware architectural regimes. An analysis of printing services in MS Windows and Solaris environments, review the age long problem of lack of device drivers for commonly (and cheaply) available Ink Jet printers for Unix (and like) operating systems. A novel method is proposed that uses agents in heterogeneous environment to overcome this problem. A new architecture that utilises Windows based printing services on a Sun Ray ultra thin client is presented to test and evaluate the proof of concept. This thesis is motivated by the need to provide a low cost printing solution to Sun Ray users. Most Windows based desktop users currently have access to variety of low cost printing solutions. Printer vendors ship device drivers only for Windows or at most Macintosh, as other operating systems such as Solaris, MVS, z/OS are used for corporate solutions and low cost desktop printing have not been a major requirement in the past.

Electronic medical records, in addition contributing to a hospital¡¦s overall electronic development and moving towards a paperless environment, also allow hospitals to share electronic medical information. To solve the problem of different medical information systems that hamper information sharing, the Executive Yuan Heath Department Commission has requested the assistance of the Taiwan Association for Medical Information to establish a standard Electronic Medical Record Template, in hopes that there is a national unified Electronic Medical Record Template that also uses the international medical information standard, so that all hospital¡¦s information is accessible, allowing for easy sharing of electronic medical records. The administration department also supports in recommending and creating related steps of incentive, to encourage professional electronic medical record researchers making up a standard internet medical information exchange mechanism, with a goal of achieving national medical information exchange. Though the Executive Yuan Heath Department Commission had a nice scheme, but the standard Electronic Medical Record Template established by Taiwan Association for Medical Information is nothing more than a Form Structure Diagram what is far from the software architecture discipline. This research aims to re-define the electronic medical record system, adopting a software architecture modeling approach. Through Structure Element Diagram, Structure Element Service Diagram, Structure Behavior Coalescence Diagram, and Sequence Diagram as four gold rules, we are able to build up any electronic medical record system architecture. Besides providing medical information personnel a more definite electronic medical record system, this research aids hospitals quickly attain an effective electronic medical record system.

The gasoline prices have risen over $130 USD per barrel since the first time oil crisis occurred in 1973. The shortages of the natural resources cause not only the prices of raw material getting higher but also hard to get. The cost of the production and marketing increased tremendously. At the present time all enterprises focus on how to get the raw material they need on producing. According to the data currently issued by U.S. Census Bureau, among the manufacturing industries, the proportion of the purchase amounts of raw material, parts, and equipments to the total cost was over 50%. The processes of the procurement are complicated and long procedures. To simplify the procurement process and save the purchasing costs become the major topics that an enterprise has to face. Unfortunately, we are lacking a set of integral model for procurement management at present. The model of system architecture is simple and concise. It describes and represents all the viewpoints simultaneously. We use to construct the mold of architecture-oriented procurement management for the references of the enterprises procurement management. This research adopting the logic approach, based on system architecture, complements with Architecture Description Language as a tool to derive the Architecture-Oriented Procurement Management Model (AOPMM). AOPMM is an integrated model of the service and interaction of all structure elements, using Six Golden Rule to express explicitly the inner knowledge. And it can be used as graphic job specifications. To face the speedily changing environment of management, AOPMM helps managers more flexible and quick to adjust the relationship inwards and outwards, to sufficiently control the resources and exchange the ideas as to make the functions of procurement management easily to appear. The results of this research can be used as the references for the enterprises procurement model in the future, to improve the efficiency of the procurement management and decision making, reduce purchasing costs, and create the higher strategic values.

Steel structure enterprises, belonging to domestic-oriented industry, are capital intensive and skill intensive as well. The ability of huge fund deployment and excellent project management is the key for such kind of enterprises to survive in the competitive environment due to the low entry barrier and little profit margin. The cost of the material accounts for the most, usually the greatest proportion of the total contract, in an entire steel structure project. Therefore, the profit in general depends heavily on the quality of material management. In this thesis, we propose Architecture Oriented Steel Structure Material Scheduling Model (AOSSMSM), to capture the current material inventory, reduce the remaining stock, improve the circulation and substitutability, and increase the profit. AOSSMSM constructs system architecture which includes enterprise architecture and software architecture, therefore highlights material¡¦s service and interaction among culture value chain in enterprise and integrates with information application system. By duplication and expansion characteristics of system architecture, AOSSMSM can be replicated effortlessly by other steel structure companies, or bridged to other models to create a new operational formation.

The electric arc furnace steelmakers produce steel across wide ranging sizes and different quality levels. The method of using electric arc for melting is an important process to produce steel from iron or scraps. The diversity of products and markets makes the steelmakers to operate within a large number of constraints of limited enterprise resources, so the electric arc furnace manufacturing system faces problems to integrate the enterprise resource planning system with the manufacturing execution system. The international automation society has come out a solution with an international standard, also known as ISA-95 aimed to help manufacturing industries such as electric arc furnace steelmakers, for the integration of enterprises and control systems. However, ISA-95 only emphasizes on the integration of enterprises and control systems, but does not provide a whole picture of the entire system. This research utilizes the tool of system architecture to develop the architecture-oriented electric arc furnace manufacturing system model, abbreviated as AOEAFMSM, to illustrate coherently the structure view and the behavior view of the electric arc furnace manufacturing system. Furthermore, AOEAFFMSM is able to fully describe many other views of the electric arc furnace manufacturing system. This capability satisfies everybody¡¦s need for capturing different views of the system. Consequentially, AOEAFFMSM simplifies the software development and reduces many communication difficulties among system developers and customers. In modeling the electric arc furnace manufacturing system model, AOEAFFMSM is superior to ISA-95.