Le présent document fait la présentation du travail réalisé au cours d'un projet de maitrise avec l'intention d'obtenir un diplôme à la maitrise en génie électrique.Le projet consistait à développer un système capable d'identifier le type de véhicule qui se retrouve dans une image infrarouge. Les deux types de véhicules visés par cette identification étaient les véhicules électriques ainsi que les véhicules à essence. Pour pouvoir être en mesure de l'identifier, il a d'abord fallu développer un système de prise d'images qui permette d'obtenir un échantillon de données comprenant des images infrarouges de différentes voitures. Deux systèmes ont été développés pour atteindre cet objectif : le premier est un système de caméra fixe et le second implique un système de mouvement des caméras. Une fois ces deux systèmes mis en place, la collecte des images fut réalisée avec chacun d'eux. Il fut remarqué que les images obtenues avec le second système étaient de meilleure qualité. Différents contextes de prise d'images ont été explorés. Lorsqu'une quantité suffisante d'images fut obtenue, l'exploration pour la sélection du réseau de neurones permettant l'identification du type de voiture a débuté. Plusieurs solutions ont été envisagées et l'évaluation de celles-ci a été réalisée. Le présent document débute par la présentation d'un article scientifique qui décrit en détails le système de mouvement des caméras / The present document elaborates on the work done for a master project with the intent to obtain a master's in electrical engineering. The project was todevelop a system that identifies the vehicle’s type within an infrared image. The two types of vehicles targeted for identification are electrical vehicles and internal combustion engine vehicles. To identify their type, the development of an acquisition system was necessary for us to collect data and build our dataset, which containsimages of each type of vehicle. Two systems were developed to achieve this goal: the first one is a fixed camera system and the second one is a motion camera system. Once each system was in place, we collected images with each system. Different situations for collecting the images were explored. It was noticed that the second system improved the images 'quality. Once we collected enough images, the exploration of the neural network for classifying the vehicle type started. Many solutions for the neural network have been presented and evaluated to choose the one that meets our needs. The present document begins with a scientific paper that has been donefor the master project which presents in details the motion camera system.

In this thesis, black-box models describing the quadcopter system dynamics for attitude control have been estimated using closed-loop data. A quadcopter is a naturally unstable multiple input multiple output (MIMO) system and is therefore an interesting platform to test and evaluate ideas in system identification and control theory on. The estimated attitude models have been shown to explain the output signals well enough during simulations to properly tune a PID controller for outdoor flight purposes. With data collected in closed loop during outdoor flights, knowledge about the controller and IMU measurements, three decoupled models have been estimated for the angles and angular rates in roll, pitch and yaw. The models for roll and pitch have been forced to have the same model structure and orders since this reflects the geometry of the quadcopter. The models have been validated by simulating the closed-loop system where they could explain the output signals well. The estimated models have then been used to design attitude controllers to stabilize the quadcopter around the hovering state. Three PID controllers have been implemented on the quadcopter and evaluated in simulation before being tested during both indoor and outdoor flights. The controllers have been shown to stabilize the quadcopter with good reference tracking. However, the performance of the pitch controller could be improved further as there have been small oscillations present that may indicate a stronger correlation between the roll and pitch channels than assumed.

There has been a significant amount of growth in soccer game attendance and media coverage since the Premier Soccer League (PSL) in South Africa was established in 1996. However, match attendance started to decline from the 2013 season. PSL teams need to do something to get the fans back into the stadiums. A stable fan base is of vital importance for a team’s competitive advantage since the growing commercialisation of sport. The South African PSL is no exception and teams need to comprehend the importance of loyal fans. Team identification is linked to higher game attendance and more fans that are loyal and, therefore, it is relevant to the PSL teams. Higher game attendance will increase the teams’ revenues through ticket sales. Highly identified fans are loyal fans and increase the likelihood of better sponsorship deals for a team. Higher levels of team identification with a team could be obtained through a unique brand personality. Brand personality is necessary for sport team managers in order to understand fans’ perception regarding a team brand. As such, managers will be able to better position the teams’ brand and reach the target market accordingly. In South Africa, members of the Generation Y cohort made up 38 percent of the country’s population in 2014 (Statistics South Africa, 2014:9), which in terms of its size, make this cohort a highly relevant market segment. Generation Y, specifically African Generation Y, represents an important current and future market segment for PSL teams. This study focuses on the student segment of the Generation Y cohort. Given that a tertiary qualification generally is associated with a higher future earning capacity and a greater role model status within a society, university students are likely to be of particular interest to marketers. The main purpose of the study was to investigate the influence of African Generation Y students’ perceived brand personality of South African PSL teams on their PSL team identification. The study used an adapted version of a team identification scale (Swanson et al., 2003) and sport brand personality scale (Braunstein & Ross, 2010) to measure the perceptions. The sampling frame for the study comprised the 26 public registered higher education institutions (HEIs) situated in South Africa. The study used one traditional university and one university of technology located in the Gauteng province. A non-probability convenience sample of 450 full time African Generation Y students was taken from the two HEIs. Of the questionnaires completed, 438 were usable. The statistical analysis of the collected data included exploratory factor analysis, descriptive statistical analysis, correlation analysis, regression analysis and independent sample t-tests. The findings suggested that Generation Y students exhibit a positive level of team identification towards their favourite team. Generation Y students perceived their favourite PSL team more hard working, confident, respected, skilled and successful. The results also suggest that the level of team identification with the PSL teams could be predicted by two dimensions (successfulness and ruggedness) of brand personality. Generation Y male and female students only differ regarding the perception of two of the brand personality dimensions (successfulness and sophistication). The findings of this study contribute to the limited literature available concerning the level of team identification and brand personality perceptions of sport teams in the South African context. The study shed light specifically on the level of team identification and brand personality perceptions of Orlando Pirates and Kaizer Chiefs. These were the top two teams identified by the respondents. These findings will assist these PSL teams to understand the brand personality perceptions of the team amongst African Generation Y fans better. The study also indicates some relationship between brand personality and team identification. The results suggest teams can use their brand personality to influence team identification amongst fans positively. These results can be used to assess and alter their current marketing strategies.

There has been a significant amount of growth in soccer game attendance and media coverage since the Premier Soccer League (PSL) in South Africa was established in 1996. However, match attendance started to decline from the 2013 season. PSL teams need to do something to get the fans back into the stadiums. A stable fan base is of vital importance for a team’s competitive advantage since the growing commercialisation of sport. The South African PSL is no exception and teams need to comprehend the importance of loyal fans. Team identification is linked to higher game attendance and more fans that are loyal and, therefore, it is relevant to the PSL teams. Higher game attendance will increase the teams’ revenues through ticket sales. Highly identified fans are loyal fans and increase the likelihood of better sponsorship deals for a team. Higher levels of team identification with a team could be obtained through a unique brand personality. Brand personality is necessary for sport team managers in order to understand fans’ perception regarding a team brand. As such, managers will be able to better position the teams’ brand and reach the target market accordingly. In South Africa, members of the Generation Y cohort made up 38 percent of the country’s population in 2014 (Statistics South Africa, 2014:9), which in terms of its size, make this cohort a highly relevant market segment. Generation Y, specifically African Generation Y, represents an important current and future market segment for PSL teams. This study focuses on the student segment of the Generation Y cohort. Given that a tertiary qualification generally is associated with a higher future earning capacity and a greater role model status within a society, university students are likely to be of particular interest to marketers. The main purpose of the study was to investigate the influence of African Generation Y students’ perceived brand personality of South African PSL teams on their PSL team identification. The study used an adapted version of a team identification scale (Swanson et al., 2003) and sport brand personality scale (Braunstein & Ross, 2010) to measure the perceptions. The sampling frame for the study comprised the 26 public registered higher education institutions (HEIs) situated in South Africa. The study used one traditional university and one university of technology located in the Gauteng province. A non-probability convenience sample of 450 full time African Generation Y students was taken from the two HEIs. Of the questionnaires completed, 438 were usable. The statistical analysis of the collected data included exploratory factor analysis, descriptive statistical analysis, correlation analysis, regression analysis and independent sample t-tests. The findings suggested that Generation Y students exhibit a positive level of team identification towards their favourite team. Generation Y students perceived their favourite PSL team more hard working, confident, respected, skilled and successful. The results also suggest that the level of team identification with the PSL teams could be predicted by two dimensions (successfulness and ruggedness) of brand personality. Generation Y male and female students only differ regarding the perception of two of the brand personality dimensions (successfulness and sophistication). The findings of this study contribute to the limited literature available concerning the level of team identification and brand personality perceptions of sport teams in the South African context. The study shed light specifically on the level of team identification and brand personality perceptions of Orlando Pirates and Kaizer Chiefs. These were the top two teams identified by the respondents. These findings will assist these PSL teams to understand the brand personality perceptions of the team amongst African Generation Y fans better. The study also indicates some relationship between brand personality and team identification. The results suggest teams can use their brand personality to influence team identification amongst fans positively. These results can be used to assess and alter their current marketing strategies.

Naval Postgraduate School's (NPS) Tactical Network Topology (TNT) experiments seek to develop, implement and identify sensor-unmanned vehicle network, and network-centric operations to assist DoD warfighters in the Global War on Terrorism (GWOT). Using biometric data for rapid identification of High Value Targets (HVT) in ground and Maritiime Interdiction Operations (MIO) is critical to the emeging special operations concept. The goal is to explore solutions and operational constraints associated with biometric data analysis and rapid identification by means of adhoc self forming sensor unmanned vehicle (UV) wireless networks. The objectives of this thesis are to look at how biometrics has performed in a testbed environment that is simulating a real special operations environment in theatre. This thesis is meant to explore and explain the biometrics process that was conducted on top of the tactical network and evaluate its performance. This thesis provided the process model for biometrics identification in the tactical networks environment. This thesis also evaluated the length of time that it took to transmit the fingerprint data from the field to the ABIS databvase, with an identification result then sent back to the field. The longest time that was observed was 70 minutes (using low bandwidth Satellite communications), while the shortest time was 4 minutes for reachback to ABIS and 2 minutes for a local database.

The topic of my thesis is a recognition in criminal proceedings. The thesis deals with recognition in terms of crimilonogy and criminal procedure. The recognition is one of the important evidence, which are often based on evidence in criminal proceedings. At first sight it may seem that a recognition is easy procedure in criminal proceedings, but closer examination reveals its complexity. The aim of the thesis is to introduce the recognition as a special way of proving and its specifics, summarization of the knowledge gained by studying textbooks and monographs, articles in journals and some judicial decisions, as well as outlining the selected questions and poiting out possible problems, especially the obligation of the defendant to attend the recognition. The thesis is divided into eight chapters, which are divided to subchapters. Following the introduction, the first chapter defines the concept of recognition, and then focuses on psychological basics of recognition, historical evolution of recognition as an evidence since 1873 and also its grounding in the Code of Criminal Procedure. Second chapter describes the relationship between recognition and other criminal methods, poiting out some of the common features and also its differences. Third chapter defines different types of recognition - with...

Cette thèse traite de l'identification des systèmes à événements discrets (SED) automatisés dans un contexte industriel. En particulier, le travail aborde les systèmes formés par un processus et un automate programmable (AP) fonctionnant en boucle fermée - l'identification a pour but d’obtenir un modèle approximatif exprimé en réseaux de Petri interprétés (RPI) à partir du comportement externe observé sous la forme d'une seule séquence de vecteurs d’entrée-sortie de l’AP. Tout d'abord, une analyse des méthodes d'identification est présentée, ainsi qu’une étude comparative des méthodes récentes pour l'identification des SED. Puis le problème abordé est décrit - des importantes caractéristiques technologiques dans les systèmes automatisés par l’AP sont détaillées. Ces caractéristiques doivent être prises en compte dans la résolution du problème, mais elles ne peuvent pas être traitées par les méthodes existantes d’identification. La contribution principale de cette thèse est la création de deux méthodes d’identification complémentaires. La première méthode permet de construire systématiquement un modèle RPI à partir d'une seule séquence entrée-sortie représentant le comportement observable du SED. Les modèles RPI décrivent en détail l’évolution des entrées et sorties pendant le fonctionnement du système. La seconde méthode considère des SED grands et complexes - elle est basée sur une approche statistique qui permettre la construction des modèles en RPI compactes et expressives. Elle est composée de deux étapes - la première calcule à partir de la séquence entrée-sortie, la partie réactive du modèle, constituée de places observables et de transitions. La deuxième étape fait la construction de la partie non-observable, en rajoutant des places pour permettre la reproduction de la séquence entrée-sortie. Les méthodes proposées, basées sur des algorithmes de complexité polynomiale, ont été implémentées en outils logiciels, lesquels ont été testés avec des séquences d’entrée-sortie obtenues à partir des systèmes réels en fonctionnement. Les outils sont décrits et leur application est illustrée à travers deux cas d’étude. / This thesis deals with the identification of automated discrete event systems (DES) operating in an industrial context. In particular the work focuses on the systems composed by a plant and a programmable logic controller (PLC) operating in a closed loop- the identification consists in obtaining an approximate model expressed in interpreted Petri nets (IPN) from the observed behaviour given under the form of a single sequence of input-output vectors of the PLC. First, an overview of previous works on identification of DES is presented as well as a comparative study of the main recent approaches on the matter. Then the addressed problem is stated- important technological characteristics of automated systems and PLC are detailed. Such characteristics must be considered in solving the identification problem, but they cannot be handled by previous identification techniques. The main contribution in this thesis is the creation of two complementary identification methods. The first method allows constructing systematically an IPN model from a single input-output sequence representing the observable behaviour of the DES. The obtained IPN models describe in detail the evolution of inputs and outputs during the system operation. The second method has been conceived for addressing large and complex industrial DES- it is based on a statistical approach yielding compact and expressive IPN models. It consists of two stages- the first one obtains, from the input-output sequence, the reactive part of the model composed by observable places and transitions. The second stage builds the non observable part of the model including places that ensure the reproduction of the observed input-output sequence. The proposed methods, based on polynomial-time algorithms, have been implemented in software tools, which have been tested with input-output sequences obtained from real systems in operation. The tools are described and their application is illustrated through two case studies.

Les techniques de mesure de champs, et en particulier celles basées sur la Corrélation d’Images Numériques (CIN), sont matures et reconnues pour la détermination de déplacements d’objets quelconques, en deux ou trois dimensions. Ces travaux ont pour objet le développement d’essais et de techniques d’identification associées, en s’appuyant sur les méthodes de mesure par CIN, pour permettre l’évaluation des caractéristiques mécaniques d’un liner en titane T35, développé pour des applications spatiales. Des essais de traction uniaxiale sont tout d'abord utilisés pour déterminer le comportement élastoplastique du T35. Plusieurs méthodes d'identification sont comparées. En particulier, il est proposé d'étendre la corrélation d'image dite "intégrée" aux cas élastoplatiques. Cette méthode permet d'évaluer à la fois les déplacements du solide observé et les paramètres mécaniques des matériaux le constituant. Elle est comparée à une méthode de recalage de référence. L'étude se porte ensuite sur l'analyse de l'amorçage des fissures dans le liner. Les chargements cycliques du réservoir imposent à la tôle une déformation biaxiale cyclique en traction-compression de quelques pourcents, qui a été reproduite expérimentalement dans une machine multiaxiale. Cette partie a notamment nécessité le développement d’un essai triaxial à rapport de charge négatif sur tôle mince. Une estimation de la limite d'amorçage a été obtenue pour les niveaux de déformation visés. Enfin, la fissuration du liner a été étudiée par des techniques d'analyse dédiées. Certaines de ces techniques sont comparées quantitativement, afin d'étudier leurs capacités et limitations respectives. Des lois de propagation de type Paris ont été identifiées directement à partir d'images du voisinage de la fissure. / Field measurement techniques, and more particularly those based on Digital Image Correlation (DIC), are mature techniques for the determination of displacements of two or three-dimensional objects. The goal of this work is to develop test and associated identification DIC-based techniques to identify the mechanical behavior of a commercially-pure titanium liner, conceived for spatial applications. Uniaxial tensile teste are first used to identify the elasto-plastic behavior of this material. Several identification methods are compared. It is especially proposed to extend said "integrated" Digital Image Correlation to elasto-plastic cases. This methods allows to identify directly the displacement of the observed solid and the mechanical parameters of the constituting material. It is compared to a reference updating method. The crack initiation is then studied in this material. Some zones of the liner will be submitted to a plastic cyclic deformation, that has to be reproduced experimentally. A dedicated test has been conceived, allowing to apply a negative load ratio to the thin sheet. An estimation of the initiation limit has been obtained for this material. Finally, the crack propagation in the liner has been studied, using dedicated techniques. Some of these have been quantitatively compared, to study their respective capabilities. Paris' propagation laws have been identified directly from a set of raw images of the crack tip vicinity.

Depuis ces dernières décennies, les matériaux composites sont de plus en plus utilisés dans l'aéronautique. Notamment, les composites tissés 3D présentent des caractéristiques matériau hors-plan intéressantes par rapport aux stratifiés. Cette technologie est développée pour les aubes FAN des moteurs d'avion. La difficulté réside dans l'identification et la validation du modèle orthotrope élastique macroscopique du pied d'aube. En effet, l'hypothèse de séparation des échelles pour l'obtention des paramètres matériau macroscopique par homogénéisation n'est pas clairement vérifiée au sein du pied d'aube comportant plusieurs zones matériau. Le composite tissé 3D formant le pied d'aube est un matériau multi-échelle complexe. Les travaux de thèse ont donc été menés afin de proposer une identification des paramètres du modèle basée sur la mesure de champs de déplacements par Corrélation d'Images Numériques (CIN) et sur la méthode d'identification de recalage de modèles éléments finis (FEMU). Cette identification a pris en compte l'influence du bruit du capteur CCD présent sur les images servant à la CIN sur l'identification des paramètres matériau. Du fait du grand nombre de paramètres matériau à identifier et des éventuels couplages entre ceux-ci, il apparaît que quelques uns ne peuvent pas être identifier à travers l'essai de traction étudié. Par conséquent, une régularisation de la FEMU a été proposée basée sur la connaissance a priori des valeurs nominales et de leur incertitude. Celle-ci consiste en une pondération intelligente vis-à-vis des données issues de l'essai afin de faire tendre les paramètres non identifiables vers leur valeur nominale. Finalement, la qualité de l'identification a été quantifiée grâce aux incertitudes sur les paramètres matériau identifiés et grâce aux cartes de résidus d'identification basées sur les images. Ces cartes traduisent la capacité du champ de déplacement calculé par le modèle éléments finis identifié à corriger l'image déformée pour la recaler sur l'image de référence, images sur lesquelles la mesure par CIN est effectuée. Ces cartes de résidus et les incertitudes obtenues permettent ainsi de valider le modèle éléments finis proposé et le cas échéant de mettre en lumière ses insuffisances. Remettre en cause les valeurs nominales ou la modélisation (par exemple le zonage matériau) pour aboutir à une description compatible avec l’expérience reste du ressort de l’ingénieur. Le travail présenté ici lui permet d’éclairer au mieux ses choix. / In recent decades, composite materials are increasingly used in aerospace. Particularly, 3D woven composite present interesting out-of-plan characteristics compared to laminates. This technology is developed for FAN blades of aircraft engines. The difficulty lies in the identification and validation of the macroscopic elastic orthotropic model of the FAN blade root. Indeed, the assumption of scales separation to obtain the macroscopic parameters by homogenization is not clearly verified in the blade root containing multiple zones. The 3D woven composite forming the blade root is a complex and multi-scale material. The thesis work has been carried out to propose an identification of model parameters based on measurement of displacement fields with Digital Image Correlation (DIC) and the identification method of finite element model updating (FEMU) . This identification has taken into account the influence of noise, coming from the CCD sensor and present on the images utilized for the DIC, on the identification of material parameters. Because of the large number of material parameters to identify and the possible couplings between them, it appears that some of them can not be identified through the tensile test studied. Therefore, a regularization was proposed based on a priori knowledge of the nominal values &#8203-&#8203-and their uncertainty. This consists in a smart balance regarding the data obtained through the test in order to force the unidentifiable parameters to their nominal value. Finally, the quality of the identification was quantified using the uncertainties on the identified material parameters and residual maps after identification based on images. These maps reflect the ability of the calculated displacement to match the corrected deformed image on the reference image, images with which the DIC is performed. These residual maps and uncertainties on material parameters allow to validate the finite element model and propose an appropriate tool to highlight its shortcomings. Challenging nominal or modeling (eg material zones) to achieve a description consistent with the experience remains the responsibility of the engineer. The work presented here allows him to enlighten its choices.

Les installations d'épuration des eaux usées sont des systèmes de grande dimension, non linéaires, sujets à des perturbations importantes en flux et en charge. Une commande prédictive (MPC) a été appliquée au Benchmark BSM1 qui est un environnement de simulation qui définit une installation d'épuration. Une identification en boucle ouverte d'une station d'épuration d'eaux usées a été réalisée pour déterminer un modèle linéaire en se basant sur un ensemble de mesures entrée sortie du Benchmark BSM1. Les réponses indicielles en boucle ouverte ont été obtenues à partir de variation échelon des entrées autour de leurs valeurs stationnaires. Le modèle tient compte des non-linéarités à travers des paramètres variables. Les réponses indicielles obtenues permettent de déterminer par optimisation les fonctions de transfert continues correspondantes. Ces fonctions de transfert peuvent être regroupées en cinq modèles mathématiques. Des fonctions de transfert continues de premier ordre, de premier ordre avec un intégrateur, des réponses inverses, de second ordre et de second ordre avec zéro représentant les réponses indicielles ont été identifiées. Les valeurs numériques des coefficients de chaque modèle choisi ont été calculées par un critère des moindres carrés. La commande prédictive (MPC) utilise le modèle obtenu comme un modèle interne pour commander le procédé. Deux stratégies de la commande prédictive DMC et QDMC d'une station d'épuration avec ou sans compensation par anticipation ont été testées. La commande par anticipation est utilisée pour réduire l'effet de deux perturbations mesurées, le débit entrant et la concentration entrante en ammonium, sur le système / Wastewater treatment processes are large scale, non linear systems, submitted to important disturbances of influent flow rate and load. Model predictive control (MPC) widely used industrial technique for advanced multivariable control, has been applied to the Benchmark Simulation Model 1 (BSM1) simulation benchmark of wastewater treatment process. An open loop identification method has been developed to determine a linear model for a set of input-output measurements of the process. All the step responses have been obtained in open loop from step variations of the manipulated inputs and measured disturbances around their steady state values. The non-linearities of the model are taken into account by variable parameters. The step responses coefficient obtained make it possible to determine by optimization the corresponding transfer functions. That functions are classified by five mathematical models, such as : first order, first order with integrator, inverse response, second order and second order with zero. The numerical values of coefficients of each model selected were calculated using a least squares criterion. Model predictive control (MPC) uses the resulting model as an internal model to control the process. Dynamic matrix control DMC and quadratic dynamic matrix control QDMC predictive control strategies, in the absence and presence of feedforward compensation, have been tested. Two measured disturbances have been used for feedforward control, the influent flow rate and ammonium concentration