Cette thèse porte sur la conception d’une classe particulière d’estimateurs d'état, les observateurs intervalles. L’objectif est d’estimer de manière garantie, les bornes supérieure et inférieure de l’ensemble admissible de l'état d’un système, à chaque instant de temps. L’approche considérée repose sur la connaissance a priori du domaine d’appartenance, supposé borné, des incertitudes du système (incertitudes de modélisation, perturbations, bruits, etc). Une classe d'observateurs intervalles à entrées inconnues est proposée pour la classe des systèmes Linéaires à Paramètres Variants (LPV). La synthèse des paramètres de l’observateur repose sur la résolution d’un problème d’optimisation sous contraintes de type inégalités matricielles linéaires (LMI) permettant de garantir simultanément les conditions d’existence de l’observateur ainsi qu’un niveau de performance, soit dans un contexte énergie, soit dans un contexte amplitude ou soit dans un contexte mixte énergie/amplitude. Plus particulièrement, la performance de l'observateur repose sur une technique de découplage pour annuler les effets des entrées inconnues et une technique d’optimisation destinée à minimiser, au sens de critères de type gain L2et/ou gain L∞, les effets des perturbations sur la largeur totale de l’enveloppe de l'état du système LPV. La méthodologie de synthèse proposée est illustrée sur un exemple académique. Enfin, la méthodologie est appliquée au cas de la phase d’atterrissage du véhicule spatial HL20, sous des conditions de simulations réalistes. / This thesis addresses the design of a class of estimator, named interval obser-ver, which evaluates in a guaranteed way, a set for the state of the system at each instant of time. The proposed approach is based on a priori knowledge of bounded sets for the system uncertainties (modeling uncertainties, disturbances, noise, etc.). A methodology to design an interval observer is proposed for the class of Linear Parameter Varying (LPV) Systems. The feasibility of the latter is based on the resolution of linear Matrix Inequalities (LMI) constraints allowing to simultaneously get the existence conditions of the intervalobserver and a certain level of a priori given performance for the state estimation of the system. Specifically, the performance of the estimates is based on a decoupling technique to avoid the effects of unknown inputs and an optimization technique to minimize, in the L2 and/or L∞ gain sense, the effects of disturbances on the estimated interval length for the state of the LPV system. The design methodology is illustrated on academic examples.Finally, the methodology is applied on the landing phase of the HL20 shuttle.

The research in this thesis is undertaken by observing that modern systems are becoming more and more complex and safety-critical due to the increasing requirements on system smartness and autonomy, and as a result health monitoring system needs to be developed to meet the requirements on system safety and reliability. The state-of-the-art approaches to monitoring system status are model based Fault Diagnosis (FD) systems, which can fuse the advantages of system physical modelling and sensors' characteristics. A number of model based FD approaches have been proposed. The conventional residual based approaches by monitoring system output estimation errors, however, may have certain limitations such as complex diagnosis logic for fault isolation, less sensitiveness to system faults and high computation load. More importantly, little attention has been paid to the problem of fault diagnosis system verification which answers the question that under what condition (i.e., level of uncertainties) a fault diagnosis system is valid. To this end, this thesis investigates the design and verification of fault diagnosis algorithms. It first highlights the differences between two popular FD approaches (i.e., residual based and fault estimation based) through a case study. On this basis, a set of uncertainty estimation algorithms are proposed to generate fault estimates according to different specifications after interpreting the FD problem as an uncertainty estimation problem. Then FD algorithm verification and threshold selection are investigated considering that there are always some mismatches between the real plant and the mathematical model used for FD observer design. Reachability analysis is drawn to evaluate the effect of uncertainties and faults such that it can be quantitatively verified under what condition a FD algorithm is valid. First the proposed fault estimation algorithms in this thesis, on the one hand, extend the existing approaches by pooling the available prior information such that performance can be enhanced, and on the other hand relax the existence condition and reduce the computation load by exploiting the reduced order observer structure. Second, the proposed framework for fault diagnosis system verification bridges the gap between academia and industry since on the one hand a given FD algorithm can be verified under what condition it is effective, and on the other hand different FD algorithms can be compared and selected for different application scenarios. It should be highlighted that although the algorithm design and verification are for fault diagnosis systems, they can also be applied for other systems such as disturbance rejection control system among many others.

La commande de systèmes physiques s’avère être une tâche difficile en général. En fonction du modèle choisi, les outils mathématiques pour l’analyse et la conception de lois de commande peuvent changés. Pour les systèmes décrits par une représentation entrée-sortie, type transfert, ou par une équation de type état, les principales informations exploitées lors de la phase d’analyse concerne la structure interne du modèle (structure finie) et la structure externe (structure à l’infini) qui permettent avant la phase de synthèse de connaître, sur le modèle en boucle ouverte, les propriétés des lois de commande envisagées ainsi que les propriétés du système piloté (stabilité…).Le travail porte principalement sur l’étude des zéros invariants des systèmes physiques représentés par bond graph, en particulier dans un contexte de modèle type LTV. L’approche algébrique est essentielle dans ce contexte car même si les aspects graphiques restent très proches du cas linéaire classique, l’extension aux modèles LTV reste très complexe d’un point de vue mathématique, en particulier pour le calcul de racines de polynômes. De nouvelles techniques d’analyse des zéros invariants utilisant conjointement l’approche bond graph (exploitation de la causalité) et l’approche algébriques ont permis de mettre en perspective certains modules associés à ces zéros invariants et de clarifier le problème d’annulation des grandeurs de sortie. L’application aux problèmes d’observateurs à entrées inconnues a permis d’illustrer nos propos sur des exemples physiques, avec certaines extensions, problèmes pour lesquels les zéros invariants apparaissent aussi comme éléments essentiels / The control synthesis of physical systems is a complex task because it requires the knowledge of a "good model" and according to the choice of a model some specific tools must be developed. These tools, mainly developed from a mathematical and theoretical point of view, must be used from the analysis step (analysis of model properties) to the control synthesis step. It is well-known that in many approaches, the properties of the controlled systems can be analyzed from the initial model. If the system is described with an input-output representation or with a state space representation, two kinds of information are often pointed out: the external structure (infinite structure) and the internal structure (finite structure). The first one is often related to the existence of some control strategies (input-output decoupling, disturbance decoupling...) and the second one gives some focus on the stability property of the controlled system.In this report, the focus has been on the study of invariant zeros of bond graph models in the context of LTV models. The algebraic approach was essential because, even if the problem is already solved for LTI bond graph models, the extension to LTV models is not so easy. The simultaneous use of algebraic and graphical approaches has been proven to be effective and convenient to solve this problem. First, some tools from the algebraic approach have been recalled in chapter one and results for the study of invariant zeros of LTI bond graph models recalled in chapter two. Some new developments are proposed in chapter three and some applications for the unknown input observer problem with some physical applications conclude this work

Orientadores: Carlos Takahiro Chone, Agricio Nubiato Crespo / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Ciências Médicas / Made available in DSpace on 2018-08-26T23:06:19Z (GMT). No. of bitstreams: 1 Alliegro_FernandoCanola_M.pdf: 11261621 bytes, checksum: 619f82b9bd6eaeef9951fee3cd054674 (MD5) Previous issue date: 2015 / Resumo: Objetivo. Avaliar a mudança no estadiamento cervical de pacientes com carcinoma espino celular (CEC) de cavidade oral e orofaringe, após alteração na padronização da análise anátomo-patológica dos linfonodos de espécimes cirúrgicos de esvaziamentos cervicais, clinicamente livres de metástases (cN0), com cortes seriados de todos os linfonodos. Materiais e Métodos. Estudo retrospectivo com 21 pacientes previamente classificados como cN0 submetidos a cirurgia para exérese do tumor primário e esvaziamento cervical eletivo. Realizou-se inicialmente a revisão das laminas originais e, posteriormente, cortes seriados com 5 µm de espessura e nova análise das lâminas. Para avaliar o impacto da mudança na sobrevida foi realizada análise de Kaplan-Meier (IC 95%). Resultados. Após a avaliação convencional, 14 pacientes foram classificados como pN0, isto é, patologicamente livres de metástases cervicais, e sete (33%) como pN+, ou seja, patologicamente acometidos. A análise com cortes seriados detectou outros três pacientes com acometimento linfonodal, previamente classificados como pN0 e 1 previamente estadiado como N1 mudou para N2b com 10 pacintes pN+ (48%). Houve aumento de 43% (3/7) no número de pacientes com acometimento linfonodal e 19% de aumento no estadiamento linfonodal após a nova padronização anátomo-patológica, porém sem impacto na sobrevida (p=0,554). Conclusão. Análise com cortes seriados de linfonodos de esvaziamento cervicais permitiu um aumento de 14% no número de pescoços com metastases ocultas e de 19% no estadiamento cervical inicial frente à técnica convencional / Abstract: Objective. Asses the change in neck staging in patients with squamous cell carcinoma (SCC) of oral cavity and oropharynx after new methodology for histopathological analysis of lymph nodes of surgical specimens from neck dissections, with step serial sections of all harvested lymph nodes, in clinically neck negative patients (cN0). It was also studied the impact of this changes on survival. Materials and methods. Retrospective study with 21 patients previously classified as cN0 undergoing surgery for excision of the primary tumor and elective neck dissection. Initially was held to revise the original pathology slides and, subsequently, step-serial sectioning with 5 µm thick and re-examination of the slides. In order to assess the impact of change in disease-free survival was held Kaplan-Meier analysis (IC 95%). Results. After conventional evaluation, 14 patients were found to be without lymphatic metastases (pN0) and seven (33%) were affected (pN+). The analysis with step serial section has detected three other patients with lymphatic metastases, previously classified as pN0, and one patient previously as N1 increased to N2b with 10 pN(+) cases (48%). There was an increase of 43% in patients with lymph node metastases and 19% of increase in neck staging after the new histopathological standardization, but without impact on survival (p = 0.554). Conclusion. Analysis of lymph nodes in surgical specimens of neck dissections with step-serial sectioning allowed an increase of 43% of necks with occult metastases and increase of neck staging in 19% when compared to conventional technique / Mestrado / Otorrinolaringologia / Mestre em Ciências Médicas

Recent developments in the computer vision field achieve state-of-the-art performance by utilizing large-scale training datasets and in the absence of that, generating synthetic datasets of said magnitude. Yet, for certain applications, it is not feasible to synthesize high fidelity training data (e.g., biomedical computer vision domain), or to achieve detailed explainability for the program's decisions. Formulating a part-based approach can help alleviate the aforementioned challenges as (i) a scene can naturally be decomposed into a hierarchical part-based structure, and (ii) using domain knowledge by incorporating the object parts' topological and geometrical constraints reduces the complexity of learning and inference, benefiting methods in terms of data efficiency and computational resources. This dissertation investigates multiple applications that benefit from a part-based solution regarding the applications' performance metrics and/or computational efficiency. We develop part-based methods for registration, segmentation, unsupervised object discovery in large-scale image collections, and unsupervised unknown foreground discovery in streaming scenarios.

Au cours de la dernière décennie, l’industrie des véhicules aériens sans pilote (UAV) a connu une croissance et une diversification immenses. De nos jours, nous trouvons des applications basées sur les drones dans un large éventail d’industries, telles que les infrastructures, l’agriculture, les transports, etc. Ce phénomène a suscité un intérêt croissant dans le domaine de la manipulation aérienne. La mise en œuvre de manipulateurs aériens dans l'industrie des UAV pourrait générer une augmentation significative du nombre d'applications possibles. Cependant, la restriction de la charge utile disponible est l’un des principaux inconvénients de cette approche. L'impossibilité d'équiper les drones de bras robotiques industriels puissants et habiles a suscité l'intérêt pour le développement de manipulateurs légers adaptés à ces applications. Dans le but de fournir une solution légère alternative aux manipulateurs aériens, cette thèse propose un bras robotique léger actionné par des fils en alliage à mémoire de forme (SMA). Bien que les fils SMA représentent une excellente alternative aux actionneurs conventionnels pour les applications légères, ils impliquent également une dynamique hautement non linéaire, ce qui les rend difficiles à contrôler. Cherchant à présenter une solution pour la tâche difficile de contrôler les fils SMA, ce travail étudie les conséquences et les avantages de la mise en œuvre des techniques de commande par retour d’état. L'objectif final de cette étude est la mise en œuvre expérimentale d'un contrôle à rétroaction d'état pour la régulation de la position du bras robotique léger proposé. Tout d'abord, un modèle mathématique basé sur un modèle physique du comportement des câbles SMA est développé et validé expérimentalement. Ce modèle décrit la dynamique du bras robotique léger proposé du point de vue de la mécatronique. Le bras robotique proposé est testé avec trois contrôleurs de retour de sortie pour le contrôle de position angulaire, à savoir un PID, un mode coulissant et une commande adaptative. Les contrôleurs sont testés dans une simulation MATLAB, puis mis en œuvre et testés expérimentalement selon différents scénarios. Ensuite, afin de réaliser la mise en œuvre expérimentale d’une technique de commande par retour d’état, un observateur d’état, à entrée inconnue, est développé. Premièrement, un modèle observable sans commutation avec une entrée inconnue est dérivé du modèle présenté précédemment. Ce modèle prend comme entrée inconnue le taux de fraction de martensite du modèle d'origine, ce qui permet d'éliminer les termes de commutation dans le modèle. Ensuite, un observateur, à entrées inconnues, basé sur le filtre de Kalman étendu et sur l’observateur à mode glissant est développé. Cet observateur permet l’estimation simultanée de l’état et des entrées inconnues. Les conditions suffisantes de convergence et de stabilité sont établies. L'observateur est testé dans une simulation MATLAB et validé expérimentalement dans différents scénarios. Enfin, une technique de commande par retour d’état est testée en simulation et implémentée de manière expérimentale pour le contrôle de position angulaire du bras robotique léger proposé. Elle est basée sur la résolution d’une équation de Riccati (SDRE). En conclusion, une analyse comparative quantitative et qualitative entre une approche de commande par retour de sortie et la une de commande par retour d’état mis en œuvre est effectuée selon plusieurs scénarios, y compris la régulation de position, le suivi de position et le suivi de charges utiles changeantes. / In the last decade, the industry of Unmanned Aerial Vehicles (UAV) has gone through immense growth and diversification. Nowadays, we find drone based applications in a wide range of industries, such as infrastructure, agriculture, transport, among others. This phenomenon has generated an increasing interest in the field of aerial manipulation. The implementation of aerial manipulators in the UAV industry could generate a significant increase in possible applications. However, the restriction on available payload is one of the main setbacks of this approach. The impossibility to equip UAVs with heavy dexterous industrial robotic arms has driven the interest in the development of lightweight manipulators suitable for these applications. In the pursuit of providing an alternative lightweight solution for the aerial manipulators, this thesis proposes a lightweight robotic arm actuated by Shape Memory Alloy (SMA) wires. Although SMA wires represent a great alternative to conventional actuators for lightweight applications, they also imply highly nonlinear dynamics, which makes them difficult to control. Seeking to present a solution for the challenging task of controlling SMA wires, this work investigates the implications and advantages of the implementation of state feedback control techniques. The final aim of this study is the experimental implementation of a state feedback control for position regulation of the proposed lightweight robotic arm. Firstly, a mathematical model based on a constitutive model of the SMA wire is developed and experimentally validated. This model describes the dynamics of the proposed lightweight robotic arm from a mechatronics perspective. The proposed robotic arm is tested with three output feedback controllers for angular position control, namely a PID, a Sliding Mode and an Adaptive Controller. The controllers are tested in a MATLAB simulation and finally implemented and experimentally tested in various different scenarios. Following, in order to perform the experimental implementation of a state feedback control technique, a state and unknown input observer is developed. First, a non-switching observable model with unknown input of the proposed robotic arm is derived from the model previously presented. This model takes the martensite fraction rate of the original model as an unknown input, making it possible to eliminate the switching terms in the model. Then, a state and unknown input observer is proposed. This observer is based on the Extended Kalman Filter (EKF) for state estimation and sliding mode approach for unknown input estimation. Sufficient conditions for stability and convergence are established. The observer is tested in a MATLAB simulation and experimentally validated in various different scenarios. Finally, a state feedback control technique is tested in simulation and experimentally implemented for angular position control of the proposed lightweight robotic arm. Specifically, continuous and discrete-time State-Dependent Riccati Equation (SDRE) control laws are derived and implemented. To conclude, a quantitative and qualitative comparative analysis between an output feedback control approach and the implemented state feedback control is carried out under multiple scenarios, including position regulation, position tracking and tracking with changing payloads.

Carcinoma of unknown primary (CUP) is an intriguing clinical finding that is defined as biopsy-proven metastasis from a malignancy in the absence of an identifiable primary site after a complete clinical work-up. CUP is a relatively common clinical entity, accounting for approximately 3–5% of all cancer diagnoses, and consists of a heterogeneous group of tumors that have acquired the capacity to metastasize before the development of a clinically evident primary lesion. Notable advances have been made over the past years in the treatment of well-defined clinical subgroups of CUP, such as women with peritoneal carcinomatosis and young adults with poorly differentiated carcinomas of midline distribution, but for the majority of patients, the prognosis still remains poor. In this review, we highlight recent advances in the diagnosis and treatment of patients with CUP syndrome, and emphasize the importance of identifying several favorable subsets of CUP, amenable to specific treatment options. In addition, we will point out novel diagnostic and therapeutic approaches which will hopefully improve both our understanding and the prognosis of this more or less neglected disease. / Unter dem Cancer of Unknown Primary (CUP)-Syndrom werden diejenigen Tumorerkrankungen zusammengefasst, bei denen auch nach Abschluss der Diagnostik nur Metastasen, jedoch kein Primärtumor gefunden wird. Das CUP-Syndrom macht ca. 3–5% aller neu diagnostizierten Malignomfälle aus und umfasst eine heterogene Gruppe von Tumoren, die die Fähigkeit zur Metastasierung erlangt haben bevor sich ein klinisch manifester Primärtumor entwickelt hat. Obwohl bemerkenswerte Fortschritte in der Behandlung von Patienten mit bestimmten, gut definierten Erkrankungssubgruppen, wie beispielsweise Frauen mit isolierter Peritonealkarzinose oder jungen Erwachsenen mit gering differenzierten Karzinomen mit Mittellinienverteilung, erzielt werden konnten, ist die Prognose bei der Mehrzahl der Patienten nach wie vor schlecht. Wir berichten im weiteren Verlauf dieser Übersichtsarbeit über Fortschritte in der Diagnostik und Therapie von Patienten mit CUPSyndrom und weisen darauf hin, dass es trotz der immer noch sehr schlechten Prognose von großer Bedeutung ist, Patienten mit bestimmten Subtypen des CUP-Syndroms zu identifizieren, die spezifischen Therapien mit der Option auf Heilung zugeführt werden sollten. Darüber hinaus möchten wir auf neuere diagnostische und therapeutische Bestrebungen aufmerksam machen, die das Verständnis und die Prognose dieses auch in der Onkologie bisher stiefmütterlich behandelten Krankheitsbildes hoffentlich verbessern werden. / Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

No / A simple method for dynamically adjusting the number of taps of linear equalisers operating in unknown channel conditions is presented. Simulations with various scenarios show that the technique successfully predicts the optimum equaliser length and is capable of adjusting it as the environment changes

No description available.

No description available.