Search

International audience; This paper deals with the design of new periodic switching control laws for high frequency DC-DC converters. The contributions are twofolds. On a first hand, the DC-DC converter model is rewritten as a periodic switched affine system thanks to a $\delta$-operator formulation, which represents an efficient framework for the numerical discretization at high frequencies. On a second hand, three different control laws are provided, the first one being the usual Lyapunov-based control law and the two others being relaxed versions of this first solution. The benefits of these two new control laws over the usual Lyapunov-based one are demonstrated on a simple example.

This paper studies linear network systems affected by multiple unknown inputs with the objective of reconstructing both the initial state and the unknown input with one time-step delay. We state conditions under which both the whole network state and the unknown input can be reconstructed from output measurements, over every window of length $N$ , with $N$ being the dimension of the system, for all system matrices that share a common zero/nonzero pattern (uniform $N$ -step strongly structural input and state observability) or at least for almost all system matrices that share a common zero/nonzero pattern (uniform $N$ -step structural input and state observability). Based on some specific assumptions on the structure of the interactions between the unknown input and the network states, we show that such a characterization depends only on strongly structural (respectively, structural) observability properties of a suitable subsystem.

International audience; Last few years have seen significant advances in the industrial application of autonomous road transport and has launched various actors in a technological race. Urban rail transport has a significant advantage in running fully automated commercial transport. Main line rail transport aims to also benefit from the advantages of automation. This work is done in the context of project TAS at IRT SystemX, with partners SNCF, Alstom Systra, and Université de Technologie de Compiègne. As part of the ongoing work this paper aims to present the context of automation for main line rail transport. The safety challenges in terms of the normative requirements, mainly for software-based solutions are discussed to identify developmental challenges. As a complement, risk analysis of some example functions for a full automation (GoA 3/4) is performed. The objective of this analysis is to identify the challenges associated with determining their safety targets, for subsequent development and demonstration to conformity.; Ces dernières années, le transport autonome a connu des avancées technologiques significatives. Le transport ferroviaire urbain présente un avantage significatif dans l'exploitation d'un transport commercial entièrement automatisé. Le transport ferroviaire de grandes lignes vise également à bénéficier des avantages de l'automatisation. Ce travail est réalisé dans le cadre du projet TAS à l'IRT SystemX, avec les partenaires SNCF, Alstom Systra et l'Université de Technologie de Compiègne. Ce papier présente le contexte de l'automatisation du transport ferroviaire pour les grandes lignes. Les défis de la démonstration de la sécurité principalement pour les logiciels applicatifs sont discutés. En complément, une analyse de risque de certaines fonctions d’une automatisation complète (GoA 3/4) est effectuée. L’objectif de cette analyse est d’identifier les défis de la détermination des objectifs de sécurité d’un tel système autonome les développements ultérieurs et l es démonstrations de la conformité.

International audience; This paper explores the different controller structures in connection with Youla-Kucera (YK) parameterization to provide stable controller reconfiguration. Mathematical stability proof is given for each of the structures. Even if stability is guaranteed, transient effects of switching between controllers can become important for the system performance. Thus, transient behavior depending on the control structure is also analysed. A comparative study is carried out among all the structures, so guidelines are given on how to choose the appropriate YK control structure.

International audience; The problem of robust stabilization with bounded feedback control is considered for a scalar reaction-diffusion system with uncertainties in the dynamics. The maximum value of the control input acting on one of the boundary points has to respect a given bound at each time instant. It is shown that, if the initial condition and the disturbance satisfy the certain bounds (computed as a function of the bound imposed on the control input), then the proposed control respects the desired saturation level and renders the closed-loop system locally input-to-state stable, that is, the trajectories with certain bound on the initial condition converge to a ball parameterized by certain norm of the disturbance.

International audience; This article deals with the stability analysis and the derivation of ISS-Lyapunov functions for infinitedimensional linear systems subject to saturations. Two cases are considered: 1) the saturation acts in the same space as the control space; 2) the saturation acts in another space, especially a Banach space. For the first case, an explicit ISS-Lyapunov function can be derived. For the second case, we prove the global asymptotic stability of the origin when considering all weak solutions.

International audience; In this paper we design and analyze distributed algorithms to compute a Nash equilibrium in potential games. Our algorithms are based on best-response dynamics, with suitable revision sequences (orders of play). We compute the average complexity over all potential games of best response dynamics under a random i.i.d. revision sequence, since it can be implemented in a distributed way using Poisson clocks. We obtain a distributed algorithm whose execution time is within a constant factor of the optimal centralized one. We then show how to take advantage of the structure of the interactions between players in a network game: non-interacting players can play simultaneously. This improves best response algorithm, both in the centralized and in the distributed case.

In this study, and through an understanding of neuronal system communication, A light was shed upon a novel model serves as an assistive technology for locked-in people suffering from Motor neuronal disease. MND patients, who have an intact minds, are able-till now-to make a mean of on-screen communication. Work was done upon the potential of brain wave activity patterns to be detected as electrical signals, classified and translated into commands following BCI constructing paradigm. However, The interface constructed was for the first time a device which can reconstruct this command physically. The intervention is in a software is used to classify imagined right and left limb movement into binary results, and then make use of the magnetic field produced by the amplified pulses to reconstruct the activity into ferrofluid droplets movement-these moved due rotation of a desk according to the data received. Objectives of the project is to address the challenges of the inaccurate performance and cost of user-training which are yet the main issues preventing BCI from being upgraded into more immersive and effective technology. The design requirements addressed were the communication speed of the droplet movement and the accuracy of hitting fixed targets. Tests were performed based on software programs accelerated changing motor imagery movement of left and right limbs. The tests achieved an average speed of 0.469 cm/s and average accuracy of 81.6% for the best volume from many were tried. A conclusion to be drawn was that the promise of this other point of view on BCI systems to be more Brain-Real World Systems.

The goal of this thesis is the investigation of Cellular Automata (CA) from the perspective of Boolean functions and combinatorial designs. Beside its theoretical interest, this research finds its motivation in cryptography, since Boolean functions and combinatorial designs are used to construct Pseudorandom Number Generators (PRNG) and Secret Sharing Schemes (SSS). The results presented in the thesis are developed along three research lines, organized as follows. The first line considers the use of heuristic optimization algorithms to search for Boolean functions with good cryptographic properties, to be used as local rules in CA-based PRNG. The main motivation is to improve Wolfram's generator based on rule 30, which has been shown to be vulnerable against two cryptanalytic attacks. The second line deals with vectorial Boolean functions induced by CA global rules. The first contribution considers the period of preimages of spatially periodic configurations in surjective CA, and analyze the cryptographic properties of CA global rules. The third line focuses on the combinatorial designs generated by CA, specifically considering Orthogonal Latin Squares (OLS), which are equivalent to SSS. In particular, an algebraic characterization of OLS generated by linear CA is given, and heuristic algorithms are used to build OLS based on nonlinear CA.; Le but de cette thèse est l'étude des Automates Cellulaires (AC) dans la perspective des fonctions booléennes et des dessins combinatoires. Au-delà de son intérêt théorique, cette recherche est motivée par ses applications à la cryptographie, puisque les fonctions booléennes et les dessins combinatoires sont utilisés pour construire des générateurs de nombres pseudo aléatoires (Pseudorandom Number Generators, PRNG) et des schémas de partage de secret (Secret Sharing Schemes, SSS). Les résultats présentés dans la thèse ont été développés sur trois lignes de recherche, organisées comme suit. La première ligne porte sur l'utilisation des algorithmes d'optimisation heuristique pour chercher des fonctions booléennes ayant des bonnes propriétés cryptographiques, à utiliser comme des règles locales dans des PRNG basés sur les AC. La motivation principale est l'amélioration du générateur de Wolfram basé sur la règle 30, qui a été montré être vulnérable vis à vis de deux attaques cryptanalytiques. La deuxième ligne s'occupe des fonctions booléennes vectorielles engendrées par les règles globales des AC. La première contribution considère la période des pré-images des configurations spatialement périodiques dans les AC surjectifs, et l'analyse des propriétés cryptographiques des règles globales des AC. La troisième ligne se concentre sur les dessins combinatoires engendrés par les AC, en considérant les Carrés Latins Orthogonaux (Orthogonal Latin Squares, OLS), qui sont équivalents aux SSS. En particulier, on donne une caractérisation algébrique des OLS engendrés par les AC linéaires, et on utilise des algorithmes heuristiques pour construire des OLS basés sur des AC non linéaires.

This thesis concerns the development of a robotic control framework for quantitative ultrasound elastography. Ultrasound elastography is a technology that unveils elastic parameters of a tissue, which are commonly related with certain pathologies. This thesis proposes three novel robotic approaches to assist examiners with elastography. The first approach deals with the control of a robot actuating an ultrasound probe to perform palpation motion required for ultrasound elastography. The elasticity of the tissue is used to design a servo control law to keep a stiff tissue of interest in the field of view of the ultrasound probe. Additionally, the orientation of the probe is controlled by a human user to explore other tissue while elastography is performed. The second approach exploits deformable image registration of ultrasound images to estimate the tissue elasticity and to help in the automatic compensation by ultrasound visual servoing of a motion introduced into the tissue. The third approach offers a methodology to feel the elasticity of the tissue by moving a virtual probe in the ultrasound image with a haptic device while the robot is performing palpation motion. Experimental results of the three robotic approaches over phantoms with tissue-like offer an excellent perspective for robotic-assistance for ultrasound elastography.; Cette thèse concerne le développement d'un cadre de contrôle robotique pour l'élastographie ultrasonore quantitative. L'élastographie ultrasonore est une technique qui dévoile les paramètres élastiques du tissu qui sont généralement liés à une pathologie. Cette thèse propose trois nouvelles approches robotiques différentes pour pour assister la procédure d'élastographie. La première approche concerne le contrôle d'un robot actionnant une sonde à ultrasons pour effectuer un mouvement de palpation nécessaire pour l'élastographie par ultrasons. L'élasticité du tissu est utilisée pour concevoir une loi d'asservissement afin de maintenir un tissu d'intérêt rigide dans le champ de vision de la sonde ultrasonore. De plus, l'orientation de la sonde est contrôlée par un utilisateur humain pour explorer différentes vues du tissu pendant que l'élastographie est effectuée. La seconde approche exploite le recalage d'images déformables avec des images ultrasonores pour estimer l'élasticité tissulaire et aider à la compensation automatique par asservissement visuel ultrasonore d'un mouvement introduit dans le tissu. La troisième approche offre une méthodologie pour ressentir l'élasticité du tissu en déplaçant une sonde virtuelle dans l'image ultrasonore avec un dispositif haptique pendant que le robot effectue un mouvement de palpation. Les résultats expérimentaux des trois approches robotiques obtenus sur des fantômes constitués de tissus démontrent l'efficacité des méthodes proposées et ouvre des perspectives intéressantes pour l'élastographie ultrasonore assistée par robot.

National audience; The emergence of mobile phones and connected objects has profoundly changed our daily lives. These devices, thanks to the multitude of sensors they embark, allow access to a broad spectrum of services. In particular, position sensors have contributed to the development of location-based services such as navigation, ridesharing, real-time congestion tracking. . . Despite the comfort offered by these services, the collection and processing of location data seriously infringe the privacy of users. In fact, these data can inform service providers about points of interests (home, workplace, sexual orientation), habits and social network of the users. In general, the protection of users’ privacy can be ensured by legal or technical provisions. While legal measures may discourage service providers and malicious individuals from infringing users’ privacy rights, the effects of such measures are only observable when the offense is already committed and detected. On the other hand, the use of privacy-enhancing technologies (PET) from the design phase of systems can reduce the success rate of attacks on the privacy of users. The main objective of this thesis is to demonstrate the viability of the usage of PET as a means of location data protection in ridesharing services. This type of location-based ser- vice, by allowing drivers to share empty seats in vehicles, helps in reducing congestion, CO2 emissions and dependence on fossil fuels. In this thesis, we study the problems of synchro- nization of itineraries and matching in the ridesharing context, with an explicit consideration of location data (origin, destination) protection constraints. The solutions proposed in this thesis combine multimodal routing algorithms with sev- eral privacy-enhancing technologies such as homomorphic encryption, private set intersection, secret sharing, secure comparison of integers. They guarantee privacy properties including anonymity, unlinkability, and data minimization. In addition, they are compared to conven- tional solutions, which do not protect privacy. Our experiments indicate that location data protection constraints can be taken into account in ridesharing services without degrading their performance.; L’émergence des téléphones mobiles et objets connectés a profondément changé notre vie quotidienne. Ces dispositifs, grâce à la multitude de capteurs qu’ils em- barquent, permettent l’accès à un large spectre de services. En particulier, les capteurs de position ont contribué au développent des services de localisation tels que la navigation, le covoiturage, le suivi de la congestion en temps réel. . . En dépit du confort offert par ces services, la collecte et le traitement des données de localisation portent de sérieuses atteintes à la vie privée des utilisateurs. En effet, ces données peuvent renseigner les fournisseurs de services sur les points d’intérêt (domicile, lieu de travail, orientation sexuelle), les habitudes ainsi que le réseau social des utilisateurs. D’une façon générale, la protection de la vie privée des utilisateurs peut être assurée par des dispositions légales ou techniques. Même si les mesures d’ordre légal peuvent dissuader les fournisseurs de services et les individus malveillants à enfreindre le droit à la vie privée des utilisateurs, les effets de telles mesures ne sont observables que lorsque l’infraction est déjà commise et détectée. En revanche, l’utilisation des technologies renforçant la protection de la vie privée (PET pour Privacy Enhancing Technologies) dès la phase de conception des systèmes permet de réduire le taux de réussite des attaques contre la vie privée des utilisateurs. L’objectif principal de cette thèse est de montrer la viabilité de l’utilisation des PET comme moyens de protection des données de localisation dans les services de covoiturage. Ce type de service de localisation, en aidant les conducteurs à partager les sièges vides dans les véhicules, contribue à réduire les problèmes de congestion, d’émissions et de dépendance aux combustibles fossiles. Dans cette thèse, nous étudions les problèmes de synchronisation d’itinéraires et d’appariement relatifs au covoiturage avec une prise en compte explicite des contraintes de protection des données de localisation (origine, destination). Les solutions proposées dans cette thèse combinent des algorithmes de calcul d’itinéraires multimodaux avec plusieurs techniques de protection de la vie privée telles que le chiffrement homomorphe, l’intersection sécurisée d’ensembles, le secret partagé, la comparaison sécurisée dentier. Elles garantissent des propriétés de protection de vie privée comprenant l’anonymat, la non-chainabilité et la minimisation des données. De plus, elles sont comparées à des so- lutions classiques, ne protégeant pas la vie privée. Nos expérimentations indiquent que les contraintes de protection des données privées peuvent être prise en compte dans les services de covoiturage sans dégrader leurs performances.

International audience; The book addresses the latest advances in research and development in the field of informatics in control, robotics and automation. With more than twenty revised and extended articles covering the theoretical aspects as well as applications and their implementation, it offers a factual and well-balanced overview of the state of the art in the field. In addition, it highlights the trends in control of intelligent robots.The book is an up-to-date source of information and inspiration for researchers, engineers and PhD students.

International audience; The problem of observer design is addressed for a class of triangular nonlinear systems with not-necessarily small delay and sampled output measurements. One more difficulty is that the system state matrix is dependent on the un-delayed output signal which is not accessible to measurement, making existing observers inapplicable. A new chain observer, composed of m elementary observers in series, is designed to compensate for output sampling and arbitrary large delays. The larger the time-delay the larger the number m. Each elementary observer includes an output predictor that is conceived to compensate for the effects of output sampling and a fractional delay. The predictors are defined by first-order ordinary differential equations (ODEs) much simpler than those of existing predictors which involve both output and state predictors. Using a small gain type analysis, sufficient conditions for the observer to be exponentially convergent are established in terms of the minimal number m of elementary observers and the maximum sampling interval.

The urban environment is strongly involved in the management of logistics flows. The pooling of these different flows provides a qualitative solution ensuring environmental, economic and societal criteria. The thesis objectives aim at supporting Urban Distribution Centers (CDU) by combining distribution logistics with reverse logistics (upstream flows and downstream flows), thereby ensuring sustainable development in modern cities. The heterogeneity of the distribution logistics practices and the reverse logistics define the core of this thesis. Sharing these different flows provides a qualitative and sustainable solution as long as it meets the aforementioned criteria (environment, economy and society). Our approach is based on the Hub Location Problem to formulate the location problem of UDC while maximizing the profit and choosing the best available location. We proceed by identifying potential available locations, while guaranteeing returned flows management (collection, sorting, resale, etc.) and reducing pollution impact. Moreover, we simulate our optimised model with simulation tools using optimization algorithms. The effectiveness of these algorithms are validated based on generated outputs (set of data) inspired by real world scenarios. Reliability is a part of this work when dealing with UDC deployment. The unavailability of the system may be due to random failures, planned preventive maintenance actions, etc. Therefore, we present a solution supporting daily demands by integrating a collaborative approach between different sites. This approach is based on inter-CDU coordination policy with a strategy of subcontracting deliveries dealing with unavailability, L’environnement urbain est fortement ancré dans la gestion des flux logistiques. La mutualisation de ces différents flux apporte alors une solution qualitative répondant aux critères environnementaux, économiques et sociétaux. Cette thèse propose une approche pour la mise en place des centres de distribution urbains (CDU) en combinant la logistique de distribution avec la logistique inverse (flux montants et flux descendants). L’objectif principal de ces CDU est d'assurer un développement durable dans les villes modernes. L’hétérogénéité des pratiques de la logistique de distribution et de la logistique inverse constitue le principal verrou scientifique. La mutualisation de ces différents flux apporte alors une solution qualitative et durable dès lors qu’elle répond aux critères attendus : environnementaux, économiques et sociétaux. Dans un contexte de logistique globale, nous nous sommes basés sur les problèmes de localisation de plateformes HLP « Hub Location Problem » afin de formuler le problème de localisation des CDU selon un critère de maximisation du profit par le choix du meilleur emplacement parmi ceux disponibles. Nous avons calculé les localisations partiellement possibles, tout en garantissant le traitement des flux retournés (collecte, tri, revente) et aussi en tenant compte des coûts socio-environnementaux. Par la suite, nous avons utilisé des outils de résolution du modèle proposé via des algorithmes d’optimisation. L’efficacité de ces algorithmes a été confirmée par un ensemble de jeux de données et de rapports d’analyse d’impacts. La dynamique liée à la fiabilité des CDU dans l’exécution du service par la prise en compte d’une politique de coordination inter-CDU fait aussi partie de notre étude. Nous nous sommes intéressés à l’aspect de disponibilité pour garantir les livraisons à temps. En effet, l’indisponibilité du système peut être due aux pannes aléatoires, aux actions de maintenance préventive planifiées, etc. Par conséquent, nous présentons une solution qui permet de satisfaire la demande journalière de manière ponctuelle par l’intégration d’une approche collaborative entre les différents sites. Il s’agit d’un modèle d’optimisation de la politique de coordination inter-CDU avec une stratégie de sous-traitance des livraisons en cas d’indisponibilité

International audience; The Internet of Things (IoT) refers to the domain of physical/logical objects connected to the Internet that can be accessible anywhere and anytime, giving the access policies. Web objects and services can be seen as distributed and cooperative agents that need to collaborate in order to reach advanced functionalities and also to optimize the overall quality offered to the end users. However, the number of the services that can be supplied through these devices on the Internet will increase more and more. In this context, performing the composition and selection of the best services according to quality-of-service (QoS) criteria become a complex task, especially when dealing with global service selection to satisfy global constraints of users in decentralized environments. For this purpose, we propose a distributed and optimal QoS selection approach based on the multiagent paradigm and the distributed constraint optimization problem (DCOP) formalism. Hence, we revisited a DCOP technique for developing an efficient algorithm: the DPOP4QoS implemented under a real-time Web protocol for IoT communication, the extensible messaging and presence protocol. The proposed algorithm takes into account the specificities of the service composition context and the satisfaction of the global user constraints. This paper also presents a set of experiments conducted under realistic distributed environment to evaluate the performance of the proposed algorithm.

International audience; This chapter presents a new methodology to perform health monitoring of hybrid systems under uncertainty. Hybrid systems can be represented as multi-mode systems with hybrid automata. Diagnosers are generated from these hybrid automata using a new data structure in order to monitor both the behavior and degradation of such systems. After a review of the state of the art on different existing solutions for diagnosis of hybrid systems under uncertainty, we propose to introduce the Hybrid Particle Petri Nets (HPPN) modeling framework. The main advantage of HPPN is that they take into account knowledge-based uncertainty in the system representation and uncertainty in the diagnosis process. The HPPN-based diagnoser deals with occurrences of unobservable discrete events (such as fault events) and it is robust to false observations. It also estimates the continuous state of the system by using particle filtering. A methodology is proposed to perform model-based diagnosis on hybrid systems by using the HPPN modeling framework. The system diagnosis is computed at any time from a HPPN-based diagnoser and contains all the hypotheses over its past mode trajectory. Each hypothesis is valued with a belief degree and includes discrete and continuous state estimates, as well as the set of faults that occurred on the system up to the current time. The HPPN-based methodology is demonstrated with an application on the K11 planetary rover prototype developed by NASA Ames Research Center. A hybrid model of the K11 is proposed and experimental results show that the approach is robust to real system data and constraints.

International audience; We propose and evaluate a three-dimensional (3D) face recognition approach that applies the speeded up robust feature (SURF) algorithm to the depth representation of shape index map, under real-world conditions, using only a single gallery sample for each subject. First, the 3D scans are preprocessed, then SURF is applied on the shape index map to find interest points and their descriptors. Each 3D face scan is represented by keypoints descriptors, and a large dictionary is built from all the gallery descriptors. At the recognition step, descriptors of a probe face scan are sparsely represented by the dictionary. A multitask sparse representation classification is used to determine the identity of each probe face. The feasibility of the approach that uses the SURF algorithm on the shape index map for face identification/authentication is checked through an experimental investigation conducted on Bosphorus, University of Milano Bicocca, and CASIA 3D datasets. It achieves an overall rank one recognition rate of 97.75%, 80.85%, and 95.12%, respectively, on these datasets.

Posture analysis in quiet standing is an essential element in evaluating human balance control. Many factors enhance the human control system’s ability to maintain stability, such as the visual system and base of support (feet) placement. In contrast, many neural pathologies, such as Parkinson’s disease (PD) and cerebellar disorder, disturb human stability. This paper addresses the problem of the automatic segmentation of stabilometric signals recorded under four different conditions related to vision and foot position. This is achieved for both control subjects and PD subjects. A hidden Markov model (HMM)-regression-based approach is used to carry out the segmentation between the different conditions using simple and multiple regression processes. Twenty-eight control subjects and thirty-two PD subjects participated in this study. They were asked to stand upright while recording stabilometric signals in mediolateral and anteroposterior directions under two permutations: feet apart and together with eyes open or closed. The results show high values for the correct segmentation rates, up to 98%, for the separation between the different conditions. The present findings could help clinicians better understand the motor strategies used by the patients during their orthostatic postures and may guide the rehabilitation process. The proposed method compares favorably with standard segmentation approaches. Note to Practitioners —In this paper, the problem of human balance control assessment is analyzed through the segmentation of the multidimensional time series of the center of pressure (CoP) displacement measurements during orthostatic postures of healthy and Parkinsonian subjects. The proposed model for automatic temporal segmentation is a specific statistical latent process model that assumes that the observed stabilometric sequence is governed by a sequence of hidden (unobserved) states or conditions. More specifically, the proposed approach is based on a specific multiple regression model that incorporates a hidden Markov process that governs the switching from one condition to another over time. The model is learned in an unsupervised context by maximizing the observed data log-likelihood via a dedicated expectation–maximization algorithm. We applied it on a real-world automatic CoP displacement excursion segmentation problem and assessed its performance by performing comparisons with alternative approaches, including well-known supervised static classifiers and the standard HMM. The results obtained are very encouraging and show that the proposed approach is quite competitive, though it works in an entirely unsupervised fashion and does not requires a feature extraction preprocessing step. The present findings could help clinicians to better understand the motor strategies used by the patients during their orthostatic postures and may guide the rehabilitation process.

National audience; This thesis deals with state estimation in discrete-time dynamic systems in the context of the integration of statistical and bounded error uncertainties. Motivated by the drawbacks of the interval Kalman filter (IKF) and its improvement (iIKF), we propose a filtering algorithm for linear systems subject to uncertain Gaussian noises, i.e. with the mean and covariance matrix defined by their membership to intervals. This new interval Kalman filter (UBIKF) relies on finding a punctual gain matrix minimizing an upper bound of the set of estimation error covariance matrices by respecting the bounds of the parametric uncertainties. An envelope containing all possible estimates is then determined using interval analysis. The UBIKF reduces not only the computational complexity of the set inversion of the matrices intervals appearing in the iIKF, but also the conservatism of the estimates. We then discuss different frameworks for representing incomplete or imprecise knowledge, including the cumulative distribution functions, the possibility theory and the theory of belief functions. Thanks to the last, a model in the form of a mass function for an uncertain multivariate Gaussian distribution is proposed. A box particle filter based on this theory is developed for non-linear dynamic systems in which the process noises are bounded and the measurement errors are represented by an uncertain Gaussian mass function. Finally, the UBIKF is applied to fault detection and isolation by implementing the generalized observer scheme and structural analysis. Through various examples, the capacity for detecting and isolating sensor/actuator faults of this tool is illustrated and compared to other approaches.; Cette thèse porte sur l’estimation d’état des systèmes dynamiques à temps discret dans le contexte de l’intégration d’incertitudes statistiques et à erreurs bornées. Partant du filtre de Kalman intervalle (IKF) et de son amélioration (iIKF), nous proposons un algorithme de filtrage pour des systèmes linéaires dont les bruits sont gaussiens incertains, c’est-à-dire de moyenne et matrice de covariance définies par leur appartenance à des intervalles. Ce nouveau filtre de Kalman intervalle (UBIKF) repose sur la recherche d’une matrice de gain ponctuelle minimisant une borne majorante de l’ensemble des matrices de covariance de l’erreur d’estimation en respectant les bornes des incertitudes paramétriques. Un encadrement de tous les estimés possibles est ensuite déterminé en utilisant l’analyse par intervalles. Le filtre UBIKF permet de réduire à la fois la complexité calculatoire de l’inversion ensembliste des matrices intervalles présent dans le filtre iIKF et le conservatisme des estimations. Nous abordons ensuite différents cadres permettant de représenter des connaissances incomplètes ou imprécises, y compris les fonctions de répartition, la théorie de possibilité et la théorie des fonctions de croyance. Grâce à cette dernière, un modèle sous forme d’une fonction de masse pour une distribution gaussienne multivariée incertaine est proposé. Un filtrage particulaire ensembliste basé sur cette théorie est développé pour des systèmes dynamiques non linéaires dans lesquels les bruits sur la dynamique sont bornés et les erreurs de mesure sont modélisées par une fonction de masse gaussienne incertaine. Enfin, le filtre UBIKF est utilisé pour la détection et l’isolation de défauts en mettant en oeuvre le schéma d’observateurs généralisé et l’analyse structurelle. Au travers de différents exemples, la capacité d’isolation de défauts capteurs/ actionneurs de cet outil est illustrée et comparée à d’autre approches.

Thèse en cotutelle avec l'Universidad de los Andes, Colombie; National audience; This thesis work was carried out in the framework of a co-tutelle between INSA, Toulouse, and the University of the Andes, Colombia, with financial support of Colciencias. This work is motivated by the need of the industry to detect abnormal situations in the plant startup and shutdown stages. Industrial plants involve integrated management of all the events that may cause accidents and translate into alarms. Process alarm management can be formulated as an event-based pattern recognition problem in which temporal patterns are used to characterize different typical situations, particularly at startup and shutdown stages. In this thesis, a new approach for alarm management based on a diagnosis process is proposed. Considering the alarms and the actions of the standard operating procedure as discrete events, the diagnosis step relies on situation recognition to provide the operators with relevant information about the failures inducing the alarm flow. The situation recognition is based on chronicles that characterize the situations of interest and are learned automatically. The chronicles are learned from representative event sequences obtained by simulation and given as input to an extended version of the Heuristic Chronicle Discovery Algorithm Modified (HCDAM). HCDAM has been extended in this thesis to account for expert knowledge in the form of specific temporal restrictions. A hybrid causal model of the process is used to verify the input event sequences and to explain and provide semantics to the learned chronicles. The Chronicle Based Alarm Management (CBAM) methodology proposed in this thesis involves different techniques to take the hybrid aspect and the standard operational procedures of the concerned processes into account. Compared to other approaches of alarm management, this approach uses information about the procedural actions related to the continuous variables behavior in a formal diagnosis process. Specific information is obtained in each step of the CBAM methodology, and it is summarized in three steps: 1. Step 1: Event type identification From the standard operating procedures and from the evolution of the continuous variables, this step determines the set of event types in startup and shutdown stages. 2. Step 2: Event sequence generation From the expertise and an event abstraction procedure this step determines the date of occurrence of each event type for constructing the representative event sequences. A representative event sequence is the set of event types with their dates of occurrence that can be associated to a specific scenario of the process. This step concludes verifying the representative event sequences using the hybrid causal graph. 3. Step 3: Chronicle database construction From the representative event sequences and temporal restrictions of each scenario, this step determines the chronicle database using the extended HCDAM algorithm. The proposed framework for alarm management is illustrated with two case studies representative of the petrochemical field.; Ce travail de thèse a été réalisé dans le cadre d’une thèse en co-tutelle entre l’INSA, Toulouse, et l’Université des Andes, Colombie, avec un financement de Colciencias. Ce travail est motivé par la nécessité pour l’industrie de détecter des situations anormales pendant les phases de démarrage et d’arrêt des installations. La sécurité des installations industrielles implique une gestion intégrée de tous les facteurs et événements pouvant causer des accidents. La gestion des alarmes peut être formulée comme un problème de reconnaissance de motifs événementiels dans lequel des modèles temporels sont utilisés pour caractériser différentes situations typiques, en particulier pendant les phases de démarrage et d’arrêt. Dans cette thèse une nouvelle approche de gestion des alarmes basée sur un processus de diagnostic est proposée. En supposant que les alarmes et les actions du mode opératoire standard sont des événements discrets, l’étape de diagnostic repose sur la reconnaissance de situation pour fournir aux opérateurs des informations pertinentes sur les défaillances induisant le flux d’alarmes. La reconnaissance de situation est basée sur des chroniques qui caractérisent les situations d’interdit et qui sont apprises de manière automatique. Les chroniques sont apprises à partir de séquences d’événements représentatives obtenues par simulation et constituant l’entrée d’une version étendue de l’Algorithme de Découverte de Chroniques Heuristique Modifié (HCDAM). HCDAM a été étendu dans cette thèse pour prendre en compte des connaissances expertes sous la forme de restrictions temporelles spécifiques. Un modèles hybride causal du procédé est utilisé pour vérifier les séquences d’entrée et pour expliquer et donner du sens aux chroniques apprises. La méthodologie de gestion des alarmes basée sur des chroniques CBAM (comme Chronicle Based Alarm Management ) proposée dans cette thèse fusionne différentes techniques pour tenir compte de l’aspect hybride et des procédures opérationnelles standard des processus concernés. Comparée aux autres approches de gestion d’alarmes, cette approche se caractérise par l’utilisation de l’information sur les actions procédurales liées au comportement des variables continues dans un processus formel de diagnostic. Des informations spécifiques sont obtenues à chaque étape de la méthodologie CBAM qui se résume en trois étapes : 1. Étape 1 : Identification du type d’événement à partir des procédures d’exploitation standard et de l’évolution des variables continues, cette étape détermine l’ensemble des types d’événements pendant les phases de démarrage et d’arrêt. 2. Étape 2 : Génération de séquence d’événements à partir de l’expertise et d’une procédure d’abstraction événementielle, cette étape détermine la date d’apparition de chaque type d’événement pour la construction des séquences d’événements représentatives. Une séquence d’événements représentatifs est l’ensemble des types d’événements avec leurs dates d’occurrence qui peuvent être associées à un scénario spécifique du processus. Cette étape se conclut avec la vérification des séquences d’événements représentatives à l’aide du modèle causal hybride. 3. Étape 3 : Construction de la base de chroniques à partir des séquences d’événements représentatives et des restrictions temporelles dans chaque scénario, cette étape détermine la base de chroniques à l’aide de l’algorithme HCDAM. La methode proposée pour la gestion des alarmes est illustrée par deux cas d’etude representatifs du domaine pétrochimique.

International audience; In the present paper, a method based on a new concept called power fuzzy soft set is proposed for multi-observer decision making problems under uncertain information. The new method applies a weighted conjunctive operator to aggregate these sets into a reliable resultant power fuzzy soft set from the input data set. To decide among the alternatives, a new ranking algorithm is introduced. The effectiveness and feasibility of this method are demonstrated by comparing it to algorithms based on the maximum score in decision making.

International audience; In recent years, a plethora of different studies for design of traditional ensemble classifiers has been proposed in order to improve final recognition accuracy. However, among the ensemble classifiers, combination methods are focused on building independent classifiers of the same or different algorithms using majority voting methods. In this paper, we present a new fusion scheme for ensemble classifiers based on a new concept called Generalized Fuzzy Soft Set (GFSS), which we apply in activity classification. Essentially, we apply a weighted aggregate operator to the output of each classifier in order to fuse the GFSS into a more reliable classifier. The proposed fusion method is based on a new ranking algorithm to classify activities. We show that the proposed method produces more accurate results than the best single classifier and its effectiveness is demonstrated by comparing it with single classifier in terms of activity recognition accuracy.

International audience; Contemporary theories assume that semantic cognition emerges from a neural architecture in which different component processes are combined to produce aspects of conceptual thought and behaviour. In addition to the state-level, momentary variation in brain connectivity, individuals may also differ in their propensity to generate particular configurations of such components, and these trait-level differences may relate to individual differences in semantic cognition. We tested this view by exploring how variation in intrinsic brain functional connectivity between semantic nodes in fMRI was related to performance on a battery of semantic tasks in 154 healthy participants. Through simultaneous decomposition of brain functional connectivity and semantic task performance, we identified distinct components of semantic cognition at rest. In a subsequent validation step, these data-driven components demonstrated explanatory power for neural responses in an fMRI-based semantic localiser task and variation in self-generated thoughts during the resting-state scan. Our findings showed that good performance on harder semantic tasks was associated with relative segregation at rest between frontal brain regions implicated in controlled semantic retrieval and the default mode network. Poor performance on easier tasks was linked to greater coupling between the same frontal regions and the anterior temporal lobe; a pattern associated with deliberate, verbal thematic thoughts at rest. We also identified components that related to qualities of semantic cognition: relatively good performance on pictorial semantic tasks was associated with greater separation of angular gyrus from frontal control sites and greater integration with posterior cingulate and anterior temporal cortex. In contrast, good speech production was linked to the separation of angular gyrus, posterior cingulate and temporal lobe regions. Together these data show that quantitative and qualitative variation in semantic cognition across individuals emerges from variations in the interaction of nodes within distinct functional brain networks.

International audience; In order to alleviate the negative effect of curse of dimensionality, band selection is a crucial step for hyperspectral image (HSI) processing. In this letter, we propose a novel unsupervised band selection approach to reduce the dimensionality for hyperspectral imagery. In order to obtain the most representative bands, the correlation matrix computed from the original HSI is used to describe the correlation characteristics among bands, while the block-diagonal structure is measured to segment all bands into a series of subspace. After applying the spectral clustering algorithm, the optimal combination of band is finally selected. To verify the effectiveness and superiority of the proposed band selection method, experiments have been conducted on three widely used real-world hyperspectral data. The results have shown that the proposed method outperforms other methods in HSI classification application.

In this paper, an active impedance control strategy for a knee-joint orthosis is proposed to assist individuals suffering from lower-limb muscular weaknesses during the swing phase of walking activities. The goal of the proposed strategy is to decrease the human effort required for ensuring a successful knee joint movement during walking without sacrificing the wearer's control priority. In this study, a gait-phase based desired knee-joint admittance model is designed by analyzing the kinematic and kinetic characteristics of the wearer's shank-foot segment during walking. Moreover, the mechanical impedance of the human/orthosis system is adapted to the desired one using an active impedance compensation. The control approach was implemented using a knee joint orthosis and tested with four healthy subjects. The EMG signals of the short head of the biceps femoris and the vastus medialis are used as metrics to evaluate the effectiveness of the proposed strategy. The experimental results show that the human effort can be significantly decreased when the wearers are assisted using the proposed approach.

International audience; The emergence of high-dimensional data requires the design of new optimization methods. Indeed, conventional optimization methods require improvements, hybridization, or parameter tuning in order to operate in spaces of high dimensions. In this paper, we present a new adaptive variant of a pattern search algorithm to solve global optimization problems exhibiting such a character. The proposed method has no parameters visible to the user and the default settings, determined by almost no a priori experimentation, are highly robust on the tested datasets. The algorithm is evaluated and compared with 11 state-of-the-art methods on 20 benchmark functions of 1000 dimensions from the CEC’2010 competition. The results show that this approach obtains good performances compared to the other methods tested.

International audience; This article proposes a robust PID adaptive controller for nonlinear systems with one or more degrees of freedom (DoF). The adaptive controller aims at minimizing the errors in trajectory tracking without requiring a prior modeling of the targeted nonlinear system. Furthermore, the proposed controller requires only the inputs and outputs of the system. And it is based on modified particle swarm optimization algorithm whose goal is to find the best PID parameters that optimize the execution of desired task by minimizing an objective function. The adaptation by the controller addresses two critical problems: The first problem is the instability of the control signal provided by the convergence phase of the classical PSO algorithm. This behavior adversely affects the lifetime of any actuator and, therefore, is undesirable. The second problem is the stagnation of the classical PSO algorithm after convergence at the immediately found optimal solution. The proposed adaptive PID controller is initially tested in simulation on a dynamical model of a robot manipulator evolving in the vertical plan. Which is followed by experimental tests performed on an actuated joint orthosis worn by human subjects having different morphologies. A comparative study with two other algorithms has been also conducted. Based on the obtained results, we conclude the efficiency of the proposed approach.

International audience; This paper presents a knowledge-based engineering framework for the design, deployment and running of context-aware monitoring agents that are dedicated to ambient assisted living applications. A new modeling approach of agent knowledge, that combines the advantages of both ontologies and object oriented modeling and programming, is proposed. In this approach, the agents’ logic is implemented using a micro-ontology and production rules based on the closed world assumption, called smart rules. These rules are managed using a standard reasoning system embedded in the agent core. Unlike semantic web approaches, the proposed approach rely on the closed world and unique name assumptions. These features are required for monitoring purposes in ambient intelligence and robotics domains. We present a practical work, where monitoring agents are instantiated in the user environment and their reasoning rules operate to handle the detection and confirmation of abnormal and emergency situations with respect to user’s context. These rules allow the agents to trigger appropriate actions with help of companion robot.

International audience; no abstract

An efficient covering of the search space is an important issue when dealing with metaheuristics. Sensitivity analysis methods aim at evaluating the influence of each variable of a problem on a mod...

International audience; This paper deals with the control of an active ankle foot orthosis (AAFO) to assist the gait of paretic patients. The AAFO system is driven by both, the residual human torque delivered by the muscles spanning the ankle joint and the AAFO's actuator's torque. A projection-based model reference adaptive control is proposed to assist dorsiflexion and plantar-flexion of the ankle joint during daily living walking activities. Unlike most classical model-based controllers, the proposed one does not require any prior estimation of the system's (foot-AAFO) parameters. The ankle reference trajectory was extracted from healthy subjects gait activities in a clinical environment. The input-to-state stability of the foot-AAFO system with respect to a bounded human muscular torque is proved in closed-loop based on a Lyapunov analysis. Preliminary experimental results with a healthy subject walking on a treadmill, show satisfactory results in terms of tracking performance and ankle assistance throughout the gait cycle.

International audience; This paper presents a bounded control strategy of a knee-joint exoskeleton that ensures knee flexion/extension movements during the assistance and rehabilitation processes. The control strategy is easy to implement, since it does not require any prior knowledge of the human effort. The proposed control law is based on nested saturations with the main advantage of explicitly avoiding the actuator's saturation, guaranteeing and enhancing the wearer's security. Lyapunov-based analysis is stated to prove: 1) the asymptotic stability of the shank-foot-exoskeleton in the absence of a human active effort; 2) the trajectories ultimate boundedness in the presence of parameter uncertainties; and 3) the input-to-state stability with respect to a bounded human active torque. Preliminary real-time experiments show satisfactory results in terms of assistance-as-needed and passive/resistive rehabilitation performances following desired trajectories expressing knee flexion/extension movements.

Accepted; International audience; The present study proposes a new approach for the assessment of the human balance control. This approach is based on the decomposition of the center of pressure displacement using empirical mode decomposition (EMD) that provides an effective time-frequency analysis of non-stationary signals. Twenty-eight healthy subjects performed quiet standing in four conditions—feet apart/together with respect to eyes open/closed—while recording the stabilometric signals in the anteroposterior (AP) and mediolateral (ML) directions. The EMD method decomposes each stabilometric signal into several subsignals called intrinsic mode functions (IMFs). Stabilogram-diffusion analysis technique is applied to generate the diffusion curve of each IMF signal. Each diffusion curve is modeled as a second-order system and provides representative features, such as the gain parameter. Analysis of the gain parameter shows the major effect of visual input and feet conditions on the strategy to control/stabilize the balance. Significant differences were found between young and elderly, and between women and men. In addition, the impact of feet position seems to be higher in ML direction than in AP direction.

International audience; Considering the evolution of neuroimaging in the medical field, some new emerging biometric modalities become interesting and promising candidates to recognize persons. These modalities are considered as a part of “Hidden Biometrics” which consists in using clinical measurements and medical imaging for recognition purpose. The main motivation in using hidden biometrics is the fact that system attacks may be extremely difficult to consider. This specificity highly contributes in increasing the robustness in terms of person verification and identification. In this paper, we deal with a novel non-invasive approach to recognize persons by extracting a brain signature, called “brainprint”. In particular, we explored the brain cortical regions of volumetric brain MRI (Magnetic Resonance Imaging) images, acquired from 220 healthy subjects. For each subject, four 3D cortical surfaces are considered, then transformed into 2D cortical folds maps. From the resulting textures, brainprints are constructed by extracting features using Wavelet Gabor Transform. These brainprints are considered in this work as a discriminative signature of the brain. In terms of performance evaluation, we show that an EER = 2.90 ± 0.47 is reached for verification mode. On the other hand, when dealing with identification, the proposed approach allows a recognition rate of 99.6%.

International audience; Facial ageing modelling has been an active research topic in the field of anthropology. Considering the fact that ageing is a non-uniform and a non-linear process for different face types (e.g. origins, gender etc.), dealing with a reliable face-ageing model may considerably help investigators working in some specific fields such as forensics. Unlike numerous studies dealing with forward or predictive face models, in this study, the authors propose a backward model aiming at estimating childhood face images using their corresponding adult face appearance as an input. For the proposed approach, face contour and different components are modified non-linearly, based on an estimated geometrical model. On the other hand, the face texture is estimated by mapping a reference face texture to the estimated geometrical model. This approach will show that it will be possible to `digitally' rejuvenate an adult person's face down to it being 3-4 years old. For evaluation purposes, a database has been created from 112 subjects. Results have been evaluated using both objective (face recognition system) and subjective (human perception) criteria. The most promising and interesting results will be highlighted further ahead.

International audience; In this work we contribute to development of a “Human-like Visual-Attention-based Artificial Vision” system for boosting firefighters’ awareness about the hostile environment in which they are supposed to move along. Taking advantage from artificial visual-attention, the investigated system’s conduct may be adapted to firefighter’s way of gazing by acquiring some kind of human-like artificial visual neatness supporting firefighters in interventional conditions’ evaluation or in their appraisal of the rescue conditions of people in distress dying out within the disaster. We achieve such a challenging goal by combining a statistically-founded bio-inspired saliency detection model with a Machine-Learning-based human-eye-fixation model. Hybridization of the two above-mentioned models leads to a system able to tune its parameters in order to fit human-like gazing of the inspected environment. It opens appealing perspectives in computer-aided firefighters’ assistance boosting their awareness about the hostile environment in which they are supposed to evolve. Using as well various available wildland fires images’ databases as an implementation of the investigated concept on a 6-wheeled mobile robot equipped with communication facilities, we provide experimental results showing the plausibility as well as the efficiency of the proposed system.

Biogeography-based optimization (BBO) is an evolutionary algorithm which is inspired by the migration of species between habitats. Almost 10 years have passed since the first BBO paper was published in 2008. BBO has successfully solved optimization problems in many different domains and has reached a relatively mature state. Considering the significant and expanding research on BBO and its applications, we find that the time is right to provide a 10-year anniversary review of the published literature, and also to point out some important avenues of future research. The purpose of this paper is to summarize and organize the literature related to the past 10 years of BBO research. Beginning with a foundation of basic BBO, we review the family of BBO algorithms and discuss BBO modifications, hybridizations, applications in science and engineering, and mathematical theory. Finally, the paper presents some interesting open problems and future research directions for BBO.

International audience; La recherche opérationnelle est une discipline des méthodes scientifiques utilisables pour élaborer de meilleures décisions. Elle aide à trouver une solution où l'homme n' en trouvait pas ou ne disposait d'aucune expérience. Elle aide à juger de la qualité d'une solution, mais aussi à confirmer/justifier des décisions. La recherche opérationnelle intervient dans des domaines très divers : Big Data, GPS, Google car ; finance, environnement et développe-ment durable, domaine spatial, politiques publiques, gestion des ressources hospitalières, etc. Elle favorise une culture scientifique dans la prise de décision. Face à son omnipré-sence dans notre environnement, il est intéressant de questionner son influence sur le Droit (confidentialité des données (Big Data) ; imputation d' obligation ; prise de décision ; information des utilisateurs, etc.). Que se passe-t-il lorsque l'aide à la décision devient systéma-tiquement « la » décision ? Qui est responsable des mauvais choix ? Existe-t-il des règles éthiques lorsque ces techniques ont un impact sur l'homme ? Traite-ton de la même façon des ressources matérielles et humaines ? La recherche opérationnelle ne se contente pas de suppléer l'activité humaine, elle la change.

In the recent years, there has been a growing interest in wireless sensor networks (WSN). Network's lifetime depends on energy efficiency and load balancing where connectivity is a very important factor. However, such connectivity can be lost due to the failure of some sensor nodes which creates disruptions to the network operations, lead to a reconfiguration of the network by generating energy losses, or in another case, the network mission fails. Energy conservation is a very important problem in WSN. In this paper, we propose a new solution for the connectivity problem when failure nodes are considered. The replacement of failed nodes is done in two phases: the first one is the search of redundant nodes using the clusterheads; the second phase is a restoration of connectivity. Performance evaluation of the proposed replacement approach shows that the results are globally satisfactory.

International audience; Model-driven engineering (MDE) and search-based software engineering (SBSE) are both relevant approaches to software engineering. MDE aims to raise the level of abstraction in order to cope with the complexity of software systems, while SBSE involves the application of metaheuristic search techniques to complex software engineering problems, reformulating engineering tasks as optimization problems. The purpose of this paper is to survey the relatively recent research activity lying at the interface between these two fields, an area that has come to be known as search-based model-driven engineering. We begin with an introduction to MDE, the concepts of models, of metamodels and of model transformations. We also give a brief introduction to SBSE and metaheuristics. Then, we survey the current research work centered around the combination of search-based techniques and MDE. The literature survey is accompanied by the presentation of references for further details.

This note follows our previous works on games with randomly arriving players [3] and [5]. Contrary to these two articles, here we seek a dynamic equilibrium, using the tools of piecewise deterministic control systems The resulting discrete Isaacs equation obtained is rather involved. As usual, it yields an explicit algorithm in the finite horizon, linear-quadratic case via a kind of discrete Riccati equation. The infinite horizon problem is briefly considered. It seems to be manageable only if one limits the number of players present in the game. In that case, the linear quadratic problem seems solvable via essentially the same algorithm, although we have no convergence proof, but only very convincing numerical evidence. We extend the solution to more general entry processes, and more importantly , to cases where the players may leave the game, investigating several stochastic exit mechanisms. We then consider the continuous time case, with a Poisson arrival process. While the general Isaacs equation is as involved as in the discrete time case, the linear quadratic case is simpler, and, provided again that we bound the maximum number of players allowed in the game, it yields an explicit algorithm with a convergence proof to the solution of the infinite horizon case, subject to a condition reminiscent of that found in [20]. As in the discrete time case, we examine the case where players may leave the game, investigating several possible stochastic exit mechanisms. MSC: 91A25, 91A06, 91A20, 91A23, 91A50, 91A60, 49N10, 93E03. Foreword This report is a version of the article [2] where players minimize instead of maximizing, and the linear-quadratic examples are somewhat different., On détermine les stratégies d'équilibre dans un jeu dynamique où des joueurs identiques arrivent de façon aléatoire, comme, par exemple, des congénères arrivant sur une même ressource. On considère aussi divers mécanismes de sortie aléatoire. On obtient des théorèmes d'existence et des algorithmes de calcul, plus explicites dans le cas particulier linéaire quadratique. Toute l'étude est conduite en horizon fini et en horizon infini, et en temps discret et en temps continu.Ce rapport est une version du working paper CRESE des mêmes auteurs (en économie mathématique), référence [2], mais où les joueurs minimisent au lieu de maximiser, et les exemples linéaires quadratiques sont un peu différents.

International audience; The integration of information and communication technologies into the Smart Grid (SG) will make it smarter to provide a more efficient power delivery, economically viable and safe. In fact, electrical systems should be controlled in more flexible way to manage critical situations such as the intermittency of renewable energy and the development of new consumers like electric vehicles (EVs). In this paper, we first propose an end-to-end SG architecture with all its associated actors such as main production, transmission, distribution, SG and cyber-security managers. In contrast to the literature, this architecture takes into account both the power and communication aspects of the SG as well as the relationship between all its components. Then, we focus on EVs as prosumers (producers and consumers thanks to their energy storage capacity) and how integrate them efficiently into the power grid. Hence, we propose in a first phase a Bayesian game-theory model that aims to integrate them into the SG and maintain the equilibrium between the offer and the demand, hence avoiding electricity intermittence. In a second phase, we concentrate on EVs as consumers and propose a new Bayesian game model to optimally integrate green electricity sources into this electricity network in order to charge EVs’ batteries. This aims to promote the electricity network capacity and have an eco-friendly effect on the environment by minimizing the usage of polluting energy sources. The obtained simulation results prove that our models help to efficiently integrate the EVs into the SG and use green electricity sources with more flexibility and less loss.

This thesis deals with new monitoring technologies applied to grain storage in silos. Since the first sedentary groups, cereals have been representing the main population and livestock’s staple food. Harvests are seasonal and their storage is therefore a vital necessity for the world’s population that will count 9 billion in 2050 compared to 7.3 billion today. In this problematic around monitoring of silos, insects represent the first grain postharvest pest causing considerable losses, in some cases up to 10% of the production. Several techniques and methods have been developed and implemented in order to face the danger of these pests. The most common method is the use of synthetic chemical insecticides, which have an obvious negative effect on both human health and the environment. This work is part of a cooperative program named ECOSILO, whose main objective is to design and prototype ecological alternatives to the systematic and massive use of insecticides. This is therefore a largely multidisciplinary approach, in which we were in charge to conceptualize a comprehensive monitoring system to: provide an earlier detection of infestations and put in place an automatic control of the attractive and insecticides diffusion. Our proposal is based on a logic of integrated pest management (IPM) and focuses on the prevention and early detection of possible infestations, in order to preserve stocks and avoid reaching high level of economic loss. Its fundamental originality lies in the proposition of a predictive model of the infestation development, which consists of coupling a thermal model of the silo with a biological model describing the insect’s growth. It is thus possible, with the support of additional meteorological data, to anticipate the development of insects and to initiate preventive alarms. This solution is accompanied by more specific electronic and computer monitoring developments in silos which are: A detection system at the entrance of silos; A supervision of the top of the hottest heap of grain; A monitoring of the heap of grain in its entire height. These developments are implementing multiple detection modes (temperature, humidity, acoustic probe, CO2). All these technical contributions are incorporated in the form of a comprehensive monitoring system equipped with a user interface that meets the requirements of the overall project ECOSILO. The functional checks have been performed in the laboratory and on real site. We conclude on the perspectives opened by this work.; Cette thèse traite des nouvelles technologies de surveillance appliquées au stockage des grains en silos. Les céréales représentent depuis les premiers regroupements sédentaires l’aliment principal pour l’Homme comme pour le bétail. Les récoltes sont saisonnières et leur stockage est donc une nécessité vitale pour une population mondiale qui compte actuellement 7,3 Mds d'habitants et qui en comptera 9 Mds en 2050. Dans cette problématique autour de la surveillance des silos, les insectes représentent les premiers ravageurs de grain après récolte, ce qui entraîne des pertes considérables pouvant atteindre dans certains cas jusqu’à 10% de la production. Plusieurs techniques et méthodes ont été développées et mises en pratique afin de faire face au danger que représentent ces ravageurs. La méthode la plus répandue est l’utilisation des insecticides chimiques de synthèse dont la dangerosité, impactant tant sur la santé humaine que sur l'environnement, est plus qu'évidente. Ce travail de thèse s'inscrit dans un programme coopératif, nommé ECOSILO, dont l'objectif majeur est de concevoir et de prototyper des solutions alternatives écologiques à l'utilisation systématique et massive de produits insecticides. Il s'agit donc d'une démarche largement pluridisciplinaire dans laquelle nous avons eu en charge d'imaginer un système complet de surveillance : détection au plus tôt des infestations et commande automatique de la diffusion d'attractifs puis d'insecticides. Notre proposition est fondée sur une logique de protection intégrée et privilégie la prévention et la détection précoce d'éventuelles infestations, ce afin de préserver les stocks et d'éviter d'atteindre un stade de perte économique trop élevé. Son originalité fondamentale réside dans la proposition d'un modèle prédictif du développement de l'infestation qui consiste à coupler un modèle thermique du silo avec un modèle biologique décrivant la croissance des insectes. Il est ainsi possible, avec l'appui de données complémentaires météorologiques, d'anticiper le développement des insectes et de lancer des alarmes préventives. Cette solution est accompagnée de développements électroniques et informatiques plus spécifiques de la surveillance continue des silos : un système de détection à l'entrée des silos ; une surveillance sur la partie haute du tas de grain la plus chaude ; une surveillance du tas de grains dans toute sa hauteur. Ces développements mettent en oeuvre des modes de détections multiples (température, humidité, sonde acoustique, CO2). Toutes ces contributions techniques sont intégrées sous la forme d'un système complet de surveillance équipé d'une interface utilisateurs qui répond aux exigences du projet global ECOSILO. Les vérifications fonctionnelles ont été réalisées en laboratoire et sur site réel. Nous concluons sur les perspectives ouvertes par ce travail.

International audience; An important challenge to be addressed in the domain of Internet of Things (IoT) is the development of efficient services selection algorithms for an optimal management of both energy and Quality of Service (QoS) in the context of IoT services composition. This issue becomes crucial in the case of large-scale IoT environments composed of thousands of distributed entities. In this paper, an energy-centered and QoS-aware services selection algorithm (EQSA) is proposed for IoT services composition. The proposed selection approach consists of preselecting the services offering the QoS level required for user's satisfaction using a lexicographic optimization strategy and QoS constraints relaxation technique. In order to reduce the energy consumption of a composite service without affecting the user's satisfaction, the most suitable services among the preselected ones are then selected using the concept of relative dominance of services in the sense of Pareto. The relative dominance of a candidate service depends on its energy profile and QoS attributes, and user's preferences. The proposed algorithm has been evaluated through several simulation scenarios. The obtained results show clearly the good performances of the EQSA algorithm in terms of selection time, energy efficiency, composition lifetime, and optimality and its added value in comparison with algorithms dealing separately with QoS and energy consumption.

International audience; Neurologic injuries, such as stroke, spinal cord injuries, and weaknesses of skeletal muscles with elderly people, may considerably limit the ability of this population to achieve the main daily living activities. Recently, there has been an increasing interest in the development of wearable devices, the so-called exoskeletons, to assist elderly as well as patients with limb pathologies, for movement assistance and rehabilitation. In this paper, we review and discuss the state of the art of the lower limb exoskeletons that are mainly used for physical movement assistance and rehabilitation. An overview of the commonly used actuation systems is presented. According to different case studies, a classification and comparison between different types of actuators is conducted, such as hydraulic actuators, electrical motors, series elastic actuators, and artificial pneumatic muscles. Additionally, the mainly used control strategies in lower limb exoskeletons are classified and reviewed, based on three types of human-robot interfaces: the signals collected from the human body, the interaction forces between the exoskeleton and the wearer, and the signals collected from exoskeletons. Furthermore, the performances of several typical lower limb exoskeletons are discussed, and some assessment methods and performance criteria are reviewed. Finally, a discussion of the major advances that have been made, some research directions, and future challenges are presented.

International audience; Wearable robotic systems have been a mechanism which clearly drives the motive of bringing back paraplegics back on their feet as well as executing difficult task beyond human ability. The purpose of this research study is to design and investigate the efficacy of rehabilitative walking in patients with lower limb disorders using oscillators which may commonly be referred to as central pattern generators (CPGs). In order to achieve this, a rhythmic trajectory is designed using Van der Pol oscillators. This rhythmic trajectory commensurates with the movement pattern of the hips and knees for a normal walking gait of humans. The dynamical model of a five-link biped exoskeletal device having four actuated joints is computed with regard to the wearer using Lagrangian principles in the sagittal plane. A feedback linearization control technique is therefore utilized for tracking the rhythmic trajectory to achieve a proper following of the human walking gait. Matlab/Simulink is used to validate this proposed strategy in the presence of uncertainties with a view to implementing it practically in the laboratory with human in the loop. Results show that humans with the aid of the exoskeleton device will possess the ability to track this rhythmic trajectory representing the hip and knee joint movements. The controller proved robust enough against disturbance.

The present paper deals with the control of a knee joint orthosis intended to be used for rehabilitation and assistive purposes. A model, integrating human shank and orthosis, is presented. To reduce the influence of the uncertainties in muscular torque modeling on the system control, a nonlinear observer is proposed to estimate the muscular torque developed by the wearer. Additionally, a robust terminal sliding mode control approach combined with the nonlinear observer is presented. To illustrate the effectiveness of the proposed control method, a comparison with two control methods, basic sliding mode and sliding mode with nonlinear observer, are also given. The asymptotic stability of the presented approaches and observer convergence are proved by means of a Lyapunov analysis. Furthermore, the proof of advantage of the robust terminal sliding mode control method with the nonlinear observer (improving the tracking precision and reducing the required time for eliminating external disturbances) is proposed as well. The experiment results show that the robust terminal sliding mode control approach combined with the nonlinear observer has a significant advantage with respect to the position tracking and robustness regarding the modeling identification errors and external disturbances. This paper deals with the control of a knee joint exoskeleton for rehabilitation purposes.An observer is proposed to estimate the muscular torque developed by the wearer.A robust terminal sliding mode control method combined with the observer is presented.Asymptotic stability is proved by means of a Lyapunov analysis.Experiment tests were conducted with three subjects.

International audience; no abstract

International audience; This paper presents a control approach of a lower limb exoskeleton by modulating the original impedance of the wearer’ swinging leg to reduce the muscular efforts. The proposed method ensures compensation of the damping and gravity effects of the wearer’s leg to desired levels offering though the possibility to adapt the lower limb wearer’s impedance as a function of the gait phase evolution. The performance of the proposed approach is evaluated by carrying out experiments with two healthy subjects. The EMG activities of the extensor muscles spanning the knee-joint are used as assesment criteria. The results show that the muscular activities required to perform the same flexion/extension movements are effectively reduced by providing lower desired joint damping and stiffness with exoskeleton’s assistance.