Ce manuscrit décrit la modélisation et la régulation de la température au sein d’un dispositif expérimental développé initialement pour valider les principes de conception adoptés pour une nouvelle machine de mesure de cylindricité au Laboratoire Commun de Métrologie du Laboratoire national de métrologie et d’essais – Conservatoire national des arts et métiers.L’appareil a été équipé de 19 sondes à résistance de platine raccordées à une référence nationale afin d’étudier l’influence sur son comportement des perturbations thermiques générées par des sources de chaleur internes et externes. L’investigation de l’influence de ces perturbations sur les mesures réalisées avec des capteurs de déplacement capacitifs a également été menée. Les perturbations thermiques internes simulant les puissances dissipées par les éléments de guidage mécaniques ont été générées par l’intermédiaire de trois films chauffants. Une modélisation par éléments finis du dispositif expérimental a été réalisée et les résultats numériques comparés à des résultats expérimentaux réalisés dans les mêmes conditions. Les écarts obtenus, de l’ordre de 0,1 °C, sont trop élevés pour que ce modèle soit adopté pour l’élaboration d’une régulation thermique en temps réel. Dans la suite, un modèle réduit a été développé à partir des données expérimentales à l’aide de la méthode d’identification modale (MIM). Le résidu obtenu lors de la comparaison des résultats issus de ce modèle et expérimentalement est inférieur à 0,003 °C. Finalement, une régulation thermique à mieux que le centième de degré a été mise en oeuvre en utilisant une commande prédictive associée à un filtre de Kalman. / This thesis describes the modelling and real-time regulation of the temperature inside an apparatus developed to validate the design principles of a cylindricity measurement machine at the Laboratoire Commun de Métrologie du Laboratoire national de métrologie et d’essais – Conservatoire national des arts et métiers. To study the effect of internal and external perturbations on the behaviour of the system, the apparatus is equipped with 19 platinum resistance thermometers calibrated with respect to the national standard. The effect of perturbations on the behaviour of capacitive displacement sensors has also been studied. The effect of internal perturbations generated by the mechanical guide rails was simulated using three film resistive heaters. Finite element modeling of the system temperature was carried out and the numerical results compared with experiment. The offsets about 0.1°C are too large for the model to be used for real-time temperature control. Subsequently, a reduced model wasdeveloped based on experimental data using the modal identification method (MIM). The residual obtained when its results are compared with experiment is 0.003°C. Finally, a temperature servo implemented using predictive control combined with a Kalman filter.

Si les flores contaminantes représentent la plupart du temps, dans les conserves, un risque industriel aujourd'hui maitrisé, la flore d'altération, de par sa résistance importante à la température, continue à constituer une cause de pertes économiques majeures. Pourtant cette dernière restait cependant peu caractérisée. En s'appuyant sur les travaux réalisés ces dernières années au sein de l'unité de microbiologie EMaiRIT'S du CTCPA (unité d'Expertise dans la Maitrise du Risque Industriel en Thermorésistants Sporulés du Centre Technique de la Conservation des Produits Agricoles), les principaux objectifs de cette thèse ont été (i) d'identifier et caractériser, en vue de sa maitrise ultérieure, la flore d'altération sporulante (ii) d'identifier l'origine de ces flores dans les conserveries et enfin (iii) de déterminer des moyens de maitrise.Pour cela, un état des lieux des bactéries sporulées d'altération des conserves a été effectué avec la collaboration de 122 conserveries sur plus de 10 ans en France. Cette caractérisation des espèces altérantes a permis l'élaboration d'un outil de biologie moléculaire (SporeTraQTM) afin d'identifier rapidement ces germes ou de pouvoir les détecter au sein d'une population complexe. En parallèle, l'amélioration de la connaissance de la thermorésistance de ces espèces, principale caractéristique de la flore sporulante, a été menée. A ce paramètre, il a été associé une relation avec la chimio résistance des spores. Identifiée, nous avons cherché à localiser cette flore d'altération au sein des usines à l'aide de plusieurs campagnes de prélèvements sur différents légumes. Au final, la flore spécifique du procédé de fabrication des conserves a été identifiée, caractérisée et localisée en vue d'améliorer la maitrise du risque microbien soit par une maitrise des contaminations et/ou un nettoyage plus performant (localisation au niveau d'étapes unitaires, efficacité de molécules sporicides) soir par un barème optimisé (en relation avec la thermorésistance). De plus, ce travail a été conduit au sein d'une approche bénéfice/risque représentant le futur de l'évolution des procédés agro-alimentaires associant amélioration de la qualité nutritionnelle et maintien de la maitrise sanitaire. Cette thèse s'appuie sur 5 publications de rang A. / Microbial contaminants of safety concern represent most of time, in canned food, an industrial risk which is well mastered. However, the spoilage flora, due to its high heat resistance, is responsible for major economic losses. Nevertheless, these bacteria remained poorly characterized. Based on the works realized during last 10 years within the EMaiRIT'S unit of microbiology of the CTCPA (expertise unit of the French Technical Center of the Preservation of Food, focused on Management of Industrial Risk liked to Heat Resistant Spores), the main objective of this thesis were: i) to identify and to characterize, with the aim of its later control, the spoilage spore forming bacteria florae ii) to identify the origin of these florae in canning factories and finally iii) to determine ways of control.For that purpose, a current inventory of spore forming bacteria in spoiled canned food was made with the cooperation of 122 canning factories over more than 10 years in France. This characterization of the spoilage species allowed the elaboration of a molecular biology tool (SporeTraQTM) for quick identification of these germs or their detection within a complex population. In parallel, the improvement of the knowledge about the heat resistance of these species, main characteristic of the spores, was led. In addition, the chemical resistance of spores was investigated. When identified, we tried to localize these spores on canning factories lines, with several sampling plans, on various vegetables. At the end, the specific spore forming bacteria related to the industrial canning process was identified, characterized and localized, allowing to improve the microbial risk control either by a more efficient cleaning, and through optimized process schedules. Furthermore, this work was driven within a benefic / risk approach representing the future of the food-processing evolution with improvement of the nutritional quality and the preservation of the sanitary control.This thesis leans on 5 publications of rank A.

Une nouvelle architecture robotique parallèle de télescope d'observation spatial actif a été développée préalablement à cette thèse. Afin de pouvoir améliorer le réglage optique du télescope, la structure robotique doit pouvoir être auto-étalonnée dans l'espace, à partir des informations disponibles (mesures proprioceptives, images, etc). Dans un premier temps, les hypothèses nécessaires pour assurer le bon déroulement de l'étalonnage sont analysées. Cette étude théorique, appuyée par des exemples, permet de définir des conditions nécessaires à l'étalonnage. Ces conditions permettent de déterminer les précisions nécessaires sur les paramètres du modèle et l'amplitude maximale du bruit de mesure admissible pour l'étalonnage. Avec ces valeurs, un critère d'arrêt pour les algorithmes d'étalonnage ayant une réelle signification physique peut être obtenu. De plus, une normalisation de la matrice d'identification est proposée, ce qui permet l'analyse de ses valeurs singulières pour détecter les problèmes d'identifiabilité des paramètres. Dans une deuxième partie, nous nous intéressons à la modélisation du télescope d'observation. Plusieurs modèles de déformation des articulations flexibles du télescope sont proposés, en considérant par exemple les équations de la théorie des poutres ou l'équilibre statique de la plate-forme. Ces modèles sont ensuite comparés expérimentalement par une analyse des résultats d'étalonnage photogrammétrique. Cette analyse permet aussi d'observer une déformation de la plate-forme mobile qui peut être approchée par l'intermédiaire de deux modèles. / A new parallel robotic architecture has previously been developed for an active space telescope. The telescope needs to be calibrated in order to improve its optical quality. This also has to be done automatically in space with the available information (propriocetive measures, images, etc) which is referred as self-calibration. First, the necessary hypothesis for calibration are analyzed. This theoretical study, illustrated with examples, leads to the definition of the necessary conditions for an efficient calibration. With those conditions, the required accuracy of the model parameters and the maximal measurement noise allowed for calibration can be calculated. A physically meaningful stop criterion for the identification algorithm can also be derived and a normalization of the identification Jacobian matrix is proposed, which permits the analysis of its singular values in order to detect identifiability problems. Concerning the telescope modeling, three kinematic models taking into account the telescope flexure joint behaviors are proposed, for example using the beam theory or the static equilibrium of the mobile platform. Those models are compared through an analysis of the calibration results using photogrammetry. This analysis highlights a deformation of the mobile platform of the telescope, that was considered using two models. Thus, the kinematic model of the telescope is composed of both a model for the flexure joints and a model approaching the mobile platform deformation. This model was validated with the calibration results of photogrammetry.

Master of Science / Department of Animal Sciences and Industry / Dale A. Blasi / The proposed U.S. National Animal Identification System (US-NAIS) has generated many concerns among beef cattle producers. The goal of the NAIS is to utilize 48-hour traceback in the event of an animal disease outbreak. The traceback would identify all animals that have had contact with the diseased animal, while linking an animal to its premises of origin. According to the Diffusion of Innovation theory, getting a new idea adopted, even when it has clear advantages, is often very difficult. However, by adopting innovations relatively sooner than others in their system, the theory demonstrates marked benefits for innovators and early adopters, as well as a widening of the socioeconomic gap. In this study beef producer demonstrated understanding, acceptance and adaptability to implement the US-NAIS. Participants for the study were selected in the spring of 2006 from a mailing list of cow-calf producers with more than 100 head of cows. BEEF Magazine provided the mailing list and a random sample of 1,000 producers was selected. The results show a knowledge gap between the proposed system and producer understanding. Producers were also divided on support for the proposed system. When ranking their level of support on a scale of 1 to 6, with 1 being strongly supportive and 6 being strongly opposed, 49% of producers showed some level of support and 48% showed some level of opposition, with a mean of 3.53 (s.d. 1.67). Data also highlights a lack of understanding of the regulations and implementation procedures among producers. The results of this study brought considerable insight into the current practices and perceptions of beef cattle producers, and will be used to develop educational materials to improve their understanding of this proposed program.

Master of Science / Department of Animal Sciences and Industry / Dale A. Blasi / Low-frequency (LF) radio frequency identification (RFID) transponders (n = 1,993) representing both full-duplex and half duplex air interface technologies were evaluated. Transponders representing five manufacturers and seven types were evaluated for read distance (RD), resonance frequency (RF) and voltage response (VR). LF RFID transceivers (n = 24) were also evaluated for performance and variance as tested by read distance. Transponders were sorted into four categories based on RD performance on three transceivers, “Top 25%,” “Middle 50%,” “Bottom 25%,” and “No Read.” These categories were used for evaluation of transponders and transceivers in experiments one and two, respectively. In experiment one, the mean RF of the “Top 25%” transponders were closer to 134.2 kHz (P < 0.05) within a transponder type (TT). TT and mean RD performance category interacted to affect the VR of transponders (P < 0.05); transponders with lower VR tended to have longer RD within a TT. In experiment two, sixty transponders from the “Middle 50%” were used to evaluate transceivers. Transceivers represented five manufacturers and five transceivers per manufacturer; one transceiver was eliminated from testing due to mechanical problems. There was a significant interaction (P < 0.0001) for TT and transceiver manufacturer. This indicated that transceiver performance was greatly dependant on TT being interrogated. TT and transceiver manufacturer interacted to affect RD variance (P < 0.05) demonstrating that transceiver RD will vary depending on TT being interrogated. In the final study, electromagnetic interference (EMI) was evaluated in fourteen livestock auction markets, four feedlots and five cattle abattoirs. The presence of EMI is known to impair the performance of RFID equipment. However, this phenomenon in livestock management settings has not been quantified in the scientific literature. EMI (134.2 + 25 kHz) was observed in all abattoirs. However, the extent and duration of EMI varied depending on individual abattoir. The processing, load and unload areas were evaluated in commercial feedlots. The most EMI was observed in the processing area. Finally, EMI was observed at the sale ring exit at two livestock auction markets. EMI exists in livestock management settings and may negatively impact the performance of LF RFID.

Three stocks of the Cape horse mackerel, Trachurus trachurus capensis, have been recognised in the ICSEAF convention region off southern Africa (Figure 1) on the basis of catch distribution and biological analyses. They were identified as being off Namibia (ICSEAF Divisions 1 . 3 and 1.4), off the West Coast of South Africa (Div. 1.6) and off the South Coast (Div. 2.1 and 2.2). Recently, speculation as to the accuracy of the classification of the South African populations has arisen. The aim of this study was to determine the number of stocks In Divisions 1.6, 2.1 and 2.2, using methodology which assessed the spatial and temporal nature and the phenotypic and genotypic characteristics of the species. The distribution of horse mackerel was evaluated by studying the trends in catch data and length frequency distributions obtained from the demersal industry during 1986 to 1988 and from the demersal biomass cruises in 1987 and 1988. Adult fish, found in Div. 2.1 and 2.2, probably migrated; during the months in which horse mackerel were expected to spawn, CPUE values were higher over the central Agulhas Bank than in the east of the study region. During the quiescent period, catches and numbers were higher in the latter region. Juvenile horse mackerel were found in the nursery areas utilised by pelagic fish and it is likely that they were transported northward from the Agulhas Bank. A decrease in biomass of individuals in Div. 1.5 indicated a separation between a northern and a southern population. Analyses of the phenotype, or epigenetic characters, of horse mackerel were used as a further test of stock integrity. Monthly samples were drawn from Div. 1.6, 2.1 and 2.2 during May 1988 to June, 1989. Comparisons for the values obtained from growth, length-at-50%-maturity and the season of otolith ring formation from each region showed no significant variation in the phenotype. Morphometric analysis proved inconclusive. Monthly gonadosomatic indices show that fish south of the Orange River share the same spawning season. Reports from the literature show that horse mackerel in the latter region differ from those of northern Namibia in spawning season, age-at-50%-maturity and season of otolith ring formation. The genetic structure of the populations of horse mackerel was evaluated by means of a restriction enzyme analysis of the Mitochondrial DNA of 37 fish collected from Divisions 1.4, 1.5, 1.6, 2.1 and 2.2. Two composite genotypes of horse mackerel were found; one belonging to fish in Div. 1.4 and one to fish in Div. 1.6, 2.1 and 2.2. The genetic distance between the two genotypes, 0.07, was the expected distance between two populations at a subspecies level. Fish in Div. 1.5 consisted of both genotypes and may be a region of mixing between the two populations. The interpretation of results was cautioned; prevlous work has shown that the migration of a small number of individuals between two stocks can be expected to maintain a low variation between the populations. In concluding, it was recommended that the horse mackerel be managed as two stocks, one in Div. 1.3 and 1.4 and one in Div. 1.6, 2.1 and 2.2

Avec la complexité croissante des systèmes dynamiques qui apparaissent dans l'ingénierie et d'autres domaines de la science, l'étude des systèmes de grande taille composés d'un ensemble de sous-systèmes interconnectés est devenue un important sujet d'attention dans différents domaines, tels que la robotique, les réseaux de transports, les grandes structures spatiales (panneaux solaires, antennes, télescopes,...), les bâtiments,… et a conduit à des problèmes intéressants d'analyse d'identification paramétrique, de contrôle distribué et d'optimisation. L'absence d'une définition universelle et reconnue des systèmes qu'on appelle "grands systèmes", "systèmes complexes", "systèmes interconnectés",..., témoigne de la confusion entre ces différents concepts et la difficulté de définir des limites précises pour tels systèmes. L'analyse de l'identifiabilité et de l'identification de ces systèmes nécessite le traitement de modèles numériques de grande taille, la gestion de dynamiques diverses au sein du même système et la prise en compte de contraintes structurelles (des interconnections,...). Ceci est très compliqué et très délicat à manipuler. Ainsi, ces analyses sont rarement prises en considération globalement. La simplification du problème par décomposition du grand système en sous-problèmes de complexité réduite est souvent la seule solution possible, conduisant l'automaticien à exploiter clairement la structure du système.Cette thèse présente ainsi, une approche décentralisée d'identification des systèmes de grande taille "large scale systems" composés d'un ensemble de sous-systèmes interconnectés. Cette approche est basée sur les propriétés structurelles (commandabilité, observabilité et identifiabilité) du grand système. Cette approche à caractère méthodologique est mise en œuvre sur des applications thermiques des bâtiments. L'intérêt de cette approche est montré à travers des comparaisons avec une approche globale. / With the increasing complexity of dynamical systems that appear in engineering and other fields of science, the study of large systems consisting of a set of interconnected subsystems has become an important subject of attention in various areas such as robotics, transport networks, large spacial structures (solar panels, antennas, telescopes, \ldots), buildings, … and led to interesting problems of parametric identification analysis, distributed control and optimization. The lack of a universal definition of systems called "large systems", "complex systems", "interconnected systems", ..., demonstrates the confusion between these concepts and the difficulty of defining clear boundaries for such systems. The analysis of the identifiability and identification of these systems requires processing digital models of large scale, the management of diverse dynamics within the same system and the consideration of structural constraints (interconnections, ...) . This is very complicated and very difficult to handle. Thus, these analyzes are rarely taken into consideration globally. Simplifying the problem by decomposing the large system to sub-problems is often the only possible solution. This thesis presents a decentralized approach for the identification of "large scale systems" composed of a set of interconnected subsystems. This approach is based on the structural properties (controllability, observability and identifiability) of the global system. This methodological approach is implemented on thermal applications of buildings. The advantage of this approach is demonstrated through comparisons with a global approach.

Dans le contexte des réseaux complexes, cette thèse s’inscrit dans deux axes : (1) Méthodologiede la détection de communautés et (2) Evaluation de la qualité des algorithmes de détection de communautés. Dans le premier axe, nous nous intéressons en particulier aux approches fondées sur les Leaders (sommets autour desquels s’agrègent les communautés). Premièrement, nous proposons un enrichissement de la méthodologie LICOD qui permet d’évaluer les différentes stratégies des algorithmes fondés sur les leaders, en intégrant différentes mesures dans toutes les étapes de l’algorithme. Deuxièmement, nous proposons une extension de LICOD, appelée it-LICOD. Cette extension introduit une étape d’auto-validation de l’ensemble des leaders. Les résultats expérimentaux de it-LICOD sur les réseaux réels et artificiels sont bons par rapport à LICOD et compétitifs par rapport aux autres méthodes. Troisièmement, nous proposons une mesure de centralité semi-locale, appelée TopoCent, pour remédier au problème de la non-pertinence des mesures locales et de la complexité de calcul élevée des mesures globales. Nous montrons expérimentalement que LICOD est souvent plus performant avec TopoCent qu’avec les autres mesures de centralité. Dans le deuxième axe, nous proposons deux méthodes orientées-tâche, CLE et PLE, afin d’évaluer les algorithmes de détection de communautés. Nous supposons que la qualité de la solution des algorithmes peut être estimée en les confrontant à d’autres tâches que la détection de communautés en elle-même. Dans la méthode CLE nous utilisons comme tâche la classification non-supervisée et les algorithmes sont évalués sur des graphes générés à partir des jeux de données numériques. On bénéficie dans ce cas de la disponibilité de la vérité de terrain (les regroupements) de plusieurs jeux de données numériques. En ce qui concerne la méthode PLE, la qualité des algorithmes est mesurée par rapport à leurs contributions dans une tâche de prévision de liens. L’expérimentation des méthodes CLE et PLE donne de nouveaux éclairages sur les performances des algorithmes de détection de communautés / In this thesis we focus, on one hand, on community detection in complex networks, and on the other hand, on the evaluation of community detection algorithms. In the first axis, we are particularly interested in Leaders driven community detection algorithms. First, we propose an enrichment of LICOD : a framework for building different leaders-driven algorithms. We instantiate different implementations of the provided hotspots. Second, we propose an extension of LICOD, we call it-LICOD. This extension introduces a self-validation step of all identified leaders. Experimental results of it-LICOD on real and artificial networks show that it outperform the initial LICOD approach. Obtained results are also competitive with those of other state-of-the art methods. Thirdly, we propose a semi-local centrality measure, called TopoCent, that address the problem of the irrelevance of local measures and high computational complexity of globalmeasures. We experimentally show that LICOD is often more efficient with TopoCent than with the other classical centrality measures. In the second axis, we propose two task-based community evaluation methods : CLE and PLE. We examine he hypothesis that the quality of community detection algorithms can be estimated by comparing obtained results in the context of other relevent tasks. The CLE approach, we use a data clustering task as a reference. The PLE method apply a link prediction task. We show that the experimentation of CLE and PLE methods gives new insights into the performance of community detection algorithms.

In the South African Social Security Agency (SASSA) Annual Report of 2011/2012 it is stated that there were more than 15 million social grants paid out to needy beneficiaries of which 10 927 731 were Child Support Grants. A major challenge that is continually being addressed is the management and administration of these grants. In particular, the focus is on service delivery and zero tolerance to fraud and corruption. SASSA has made various attempts to address these issues, such as the rollout of biometric smart cards in 2012. This research endeavour attempts to discover whether a framework can be designed where necessary factors are taken into consideration to provide for an efficient social grant application and delivery process that uses biometrics. The framework aims to suggest improvements in the use of biometrics for the social grants. Seeing that biometrics in this case is used as a technology to improve a system involving humans, this study followed a Design Science approach and made use of a case study to collect the data required for the study. Literature studies reviewed the fields of social grants and biometrics. The challenges and lessons learnt from current implementations of social grants and biometrics within the South African context and further abroad were also relevant for the study. The framework that resulted from the above was evaluated for validity and applicability after which a modified framework is presented. The research concludes with specific implementation guidelines as well as areas for future research.

The goal of this thesis is to develop an algorithm to process infrared images and achieve automatic identification of moving subjects with fever. The identification is based on two main features: the distinction between the geometry of a human face and other objects in the field of view of the camera, and the temperature of the radiating object. Infrared thermography is a remote sensing technique used to measure temperatures based on emitted infrared radiation. Applications include fever screening in major public places such as airports and hospitals. Current accepted practice of screening requires people to stay in a line and temperature measurements are carried out for one person at a time. However in the case of mass screening of moving people the accuracy of the measurements is still under investigation. An algorithm constituting of image processing to threshold objects based on the temperature, template matching and hypothesis testing is proposed to achieve automatic identification of fever subjects. The algorithm was first tested on training data to obtain a threshold value (used to discriminate between face and non face shapes) corresponding to a false detection rate of 5%, which in turn corresponds to 85% probability of detection using Neyman-Pearson criterion. By testing the algorithm on several simulated and experimental images (which reflect relevant scenarios characterizing crowded places) it is observed that it can be beneficially implemented to introduce automation in the process of detecting moving subjects with fever.