No description available.

No description available.

The patient's failure to cooperate effectively in the patient/physician (patient and physician) interaction has been shown to be a problem of significant magnitude. In the present study, an attempt was made to identify specific, patient behaviors which might be related to physician judgment of a good patient and progress of treatment. A checklist of 37 behaviors was compiled. A series of 100 patients was observed during their interaction with physicians and occurrences of behaviors from the checklist were noted by an experimenter. Physicians also indicated whether the patient was considered to be a good patient and whether treatment was progressing as expected. For every third patient, physicians noted the occurrence of behaviors from the checklist. An association was found between some behaviors from the checklist and the physicians' judgment. There was also shown to be a difference in the ability of the experimenter and the physicians involved to detect these behaviors.

This doctoral thesis deals with EMC pre-compliance testing of radiated disturbance. The thesis summarizes present pieces of knowledge about restricting of disturbing ambient signals and focuses on restricting of test site imperfection influence on the measurement results. Pre-compliance test site “calibration” uses comparison signal sources. There were added up the recommendations for “calibration” steps, too. The semi-anechoic vs. fully-anechoic chambers measurements results were compared. Some methods for dealing with ambient disturbing signals were discussed and high attention was applied to the test site choice and the minimal dimension geometry was recommended. There are presented recommendations for pre-compliance radiated disturbance measurements in the conclusion with the stress on available test equipment.

L'une des façons les plus efficace de s'assurer du bon fonctionnement d'un système informatique est de les représenter par des modèles mathématiques. De nombreux travaux ont été réalisés en utilisant des automates finis comme modèles, nous essayons ici d'étendre ces travaux à des modèles infinis. Dans cette thèse, nous nous intéressons à quelques problèmes dans lesquels un système est observé de façon incomplète. Dans ce cas, il est impossible d'accéder à certaines informations internes. La diagnosticabilité d'une propriété donnée consiste à vérifier qu'à l'exécution du système, un observateur sera en mesure de déterminer avec certitude que la propriété est vérifiée par le système. L'opacité consiste, réciproquement, à déterminer qu'un doute existera toujours. Une autre application concerne la génération de cas de test. Une fois encore, on considère qu'un observateur n'accède qu'à une partie des événements se produisant dans le système (en général les entrées et les sorties). À partir d'une spécification, on produit automatiquement des cas de test, qui ont pour but de détecter des non-conformités (elles même formalisées de façon précise). Ces trois problèmes ont été étudiés pour des modèles finis. Dans cette thèse, nous étendons leur étude aux modèles récursifs, pour cela nous avons introduit notre propre modèle, les RTS, qui sont une généralisation des automates à pile, et d'autres modèles de la récursivité. Nous adaptons ensuite les techniques utilisées sur des modèles finis, qui servent à résoudre les problèmes qui nous intéressent. / An effective way to ensure the proper functioning of a computer system is to represent it by using mathematical models . Many studies have been conducted using finite automata as models, in this thesis we try to extend these works to infinite models. We focus on three problems in which a system is partially observed. In this case, it is impossible to access certain internal informations. Diagnosability of a given property consist in checking, that, during the execution of the system, an observer will be able to determine with certainty that the property is verified by the system. Conversely, the opacity consists in determining if a doubt will always exist. Another application is the generation of test cases. Once again, we consider that an observer accesses only some events of the system (typically the inputs and outputs): from a specification, we automatically generate test cases, which are designed to detect non-conformance. These three problems have been studied for finite models. In this thesis, we extend their study to recursive models. For this purpose, we have introduced a new model, the RTS, which are a generalization of pushdown automata and other models of recursion. In order to solve problems of interest, we adapt the techniques used in finite models.

Les matériaux plastiques contiennent des additifs, qui ne sont pas fixés dans la matrice polymère et risquent migrer dans les aliments. La directive européenne 2002/72 a introduit la possibilité de démontrer l’aptitude au contact alimentaire de ces matériaux à partir d’approches prédictives, dont l’application est limitée par la disponibilité de données de formulation et de physico-chimie. Ces travaux visent à adapter et développer des approches analytiques rapides pour identifier et quantifier les substances majoritaires contenues dans les plastiques et à développer une approche générique de prédiction des coefficients de partage entre des polymères et les simulants de l’aliment. Des méthodes conventionnelles d’extraction par solvant et de quantification en CLHP-UV-DEDL et CPG-DIF ont été comparées pour quatre formulations modèles de PEHD et PS. Une méthode rapide de déconvolution de spectres infrarouge d’extraits de PEHD a été développée pour identifier et quantifier les additifs. Un modèle prédictif des coefficients d’activité dans les PE et les simulants est proposé. Les contributions enthalpique et entropique non configurationnelle sont évaluées à partir d’un échantillonnage des énergies de contact paire à paire. Il est démontré que la contribution entropique configurationnelle est indispensable à la description de l’affinité de molécules de grande taille dans les simulants polaires ou non constitués de petites molécules. Des arbres de décision combinant approche expérimentale et modèle sont finalement discutés dans la logique de démonstration de la conformité et de veille sanitaire / Plastic packagings are formulated with additives, which can migrate from materials into foodstuffs. According to European directive 2002/72/EC, the ability of plastic materials to be used in contact with food can be demonstrated using modelling tools. Their use is however limited due to availability of some data, like the formulation of materials and partition coefficients of substances between plastics and food. On the one hand this work aims to develop the ability of laboratories to identify and quantify the main substances in plastic materials, and on the other hand it aims to develop a new method to predict partition coefficients between polymers and food simulants. Four formulations of both HDPE and PS were chosen and used during the work. Standard extraction methods and quantification methods using HPLC-UV-ELSD and GC-FID were compared. A new deconvolution process applied on infrared spectra of extracts was developed to identify and quantify additives contained in HDPE. Activity coefficients in both phases were approximated through a generalized off-lattice Flory-Huggins formulation applied to plastic materials and to liquids simulating food products. Potential contact energies were calculated with an atomistic semi-empirical forcefield. The simulations demonstrated that plastic additives have a significant chemical affinity, related to the significant contribution of the positional entropy, for liquids consisting in small molecules. Finally, decision trees, which combine both experimental and modelling approaches to demonstrate the compliance of plastic materials, were discussed

This thesis focuses on design and implementation of a testing framework for different implementation types of RISC-V architecture. It describes history, instruction set and processor modes which are supported by this architecture. Further, the current methodologies and testing frameworks implemented in Python are discussed. Emphasis is placed on the analysis of compliance tests. In the practical part, the design and implementation of a framework for execution of compliance tests for models, which can be implemented in various ways, either as an ISA simulator or a hardware model, is done. The secondary aim of the thesis is to create a graphical user interface for quick and easy test configuration. Finally, the results are evaluated and the possibilities of further development are discussed.