Submitted by Geandra Rodrigues (geandrar@gmail.com) on 2018-07-19T15:32:30Z No. of bitstreams: 1 felipegomesduque.pdf: 3507645 bytes, checksum: d7e6e84ea872f86ac2bd64caa361fce0 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2018-07-23T15:26:16Z (GMT) No. of bitstreams: 1 felipegomesduque.pdf: 3507645 bytes, checksum: d7e6e84ea872f86ac2bd64caa361fce0 (MD5) / Made available in DSpace on 2018-07-23T15:26:16Z (GMT). No. of bitstreams: 1 felipegomesduque.pdf: 3507645 bytes, checksum: d7e6e84ea872f86ac2bd64caa361fce0 (MD5) Previous issue date: 2018-04-12 / O presente trabalho propõe uma metodologia de planejamento de medição em Sistemas de Distribuição de Energia Elétrica (SDE) e um novo método para estimação de estados destes sistemas. Para tanto, a técnica metaheurística de otimização bio-inspirada denominada Modified Monkey Search (MMS) é proposta para alocação ótima de medidores inteligentes e unidades de medição fasorial. O modelo de otimização é multiobjetivo e visa a maximização da eficácia do processo de estimação de estados com o custo mínimo de investimento em sistemas de medição. O método de Pareto é associado ao algoritmo MMS para o tratamento adequado destes objetivos conflitantes considerando-se custos reais associados aos equipamentos de medição. Adicionalmente, um novo método de estimação de estados baseado na modelagem de um Fluxo de Potência Ótimo (FPO) modificado é proposto, cuja resolução é dada pelo Método de Pontos Interiores (MPI). O algoritmo MMS determina as variáveis discretas associadas aos tipos de equipamentos de medição, bem como aos locais de instalação dos mesmos no SDE. Estudos são realizados para comparar a nova metodologia de estimação de estados proposta com uma metodologia tradicional, bem como para comparar os resultados da metaheurística de otimização aplicada ao problema com outras técnicas desenvolvidas para esta finalidade. Os estudos são conduzidos com sistemas da literatura, além de um sistema real de médio porte de uma concessionária brasileira. / The present work proposes an approach for planning the measurement locations in Electric Distribution Systems (EDS) and a new method for static state estimation. The bio-inspired meta-heuristic optimization technique called Modified Monkey Search (MMS) is proposed for optimal allocation of smart meters and phasor measurement units. The optimization model is multiobjective and aims at maximizing the efficiency of the state estimation process with minimum measurement investment costs. The Pareto’s method is associated with the MMS algorithm for handling the conflicting objectives in a suitable manner by considering real costs related to measurement equipments. In addition, a new method for static state estimation based on the modeling of a modified Optimal Power Flow (OPF) is proposed, whose solution is given by the Interior Point Method (IPM). The MMS algorithm determines the discrete variables related to types and location of measurement equipments in the system. Studies are made to compare the new approach for static state estimation with a traditional method, as well as to compare the results from the meta-heuristic optimization applied to the problem with existing techniques. The studies are performed using systems from the literature, as will as a practical medium size distribution network from a Brazilian utility.

Le but de ce travail de doctorat est de contribuer à l'optimisation de l'accumulation de lipides par les levures oléagineuses, et plus particulièrement par la levure Yarrowia lipolytica à partir du glucose, avec une stratégie d’optimisation dynamique à partir d’une commande basée sur un modèle. L’étude bibliographique permet de faire le bilan des connaissances antérieures pour identifier les différentes méthodologies existantes pour l’optimisation des procédés en structurant trois parties principales: la modélisation, la commande, et le suivi des procédés. Dans ce contexte, cinq modèles ont été proposés pour décrire l'accumulation de lipides. Le premier est un modèle non structuré basé sur des cinétiques de Monod et d’inhibition. Le deuxième s’appuie sur le modèle de Droop (quota) précédemment utilisé pour décrire l’accumulation de lipides dans les microalgues. Les trois derniers sont des modèles métaboliques dynamiques qui combinent les cinétiques avec un réseau métabolique réduit, obtenu à partir des modes élémentaires. Les cinq modèles ont été calibrés et validés en utilisant plusieurs jeux des données expérimentales. Néanmoins, un des avantages des modèles métaboliques dynamiques présentés est la possibilité de décrire les basculements métaboliques. Deux stratégies de commande multi-objective visant à maximiser la productivité des lipides et la fraction en teneur lipidique ont été proposées. Dans la première, les deux objectifs ont été pondérés par le calcul statique des fronts de Pareto, et intégrés à la stratégie de commande avec un modèle dynamique métabolique. La deuxième stratégie est basée sur des fonctions linéaires par morceaux en intégrant le modèle quota. Les simulations de la commande montrent la possibilité d’atteindre des teneurs en lipides entre 0,21 – 0,26 gLIP.gX-1 et productivités entre 0,78 – 1,02 g.(L-h)-1 en diminuant le temps de la culture à 20 h. Des capteurs logiciels ont été proposés afin de pallier le manque de capteurs en ligne en corrélant des mesures en ligne (i.e. pO2 et la base ajoutée pour le pH) par des algorithmes de types machines à vecteurs supports. La validation expérimentale des stratégies de commande est la principale perspective de ce travail. / This PhD thesis aims at optimizing lipid accumulation by oleaginous yeast, and most particularly by the yeast Yarrowia lipolytica from glucose. This optimization is addressed from a mathemantical point of view based on automatic control laws, where model-based control strategies are proposed. The bibliographic review compiles and evaluates previous works to identify the different existing methodologies to attain the optimization, which is divided in three main axes: modeling, control strategies, and monitoring. In this context, five different models are proposed to describe lipid accumulation. The first model is based on Monod and inhibition kinetics (unstructured), and the second on the Droop quota model (quota) previously used for microalgae. The last three are dynamic metabolic models that combine kinetics with metabolic models based on the stoichiometry of metabolism. These three models used a reduced metabolic network decomposed into elementary flux modes. The five models were successfully calibrated and validated with different experimental data. Nonetheless, the dynamic metabolic models presented highlighting features such as the description of metabolic shifts. Two approaches of multi-objective control strategies aiming at maximizing lipid productivity and lipid content fraction were proposed. In the first, the two objectives were weighted by static calculation of Pareto fronts, and integrated to the control strategy by dynamic optimization algorithms with a dynamic metabolic model. The second strategy used a constant weighed objective function solved by piecewise linear functions by integrating the quota model. The simulation results of the optimization attained lipid contents between 0.21 – 0.26 gLIP.gX-1 and productivities between 0.78 – 1.02 g.(L-h)-1 shortening the culture time to 20 h. Soft-sensors were developed by correlating on-line measurements (i.e. pO2 and the added base for pH) through support vector machines in order to overcome the lack of measurements. The perspective is to experimentally validate the control strategies.

Monoclonal antibodies (MAbs) have an expanding market for use in diagnostic and therapeutic applications. Industrial production of these biopharmaceuticals is usually achieved based on fed-batch cultures of mammalian cells in bioreactors (Chinese hamster ovary (CHO) and Hybridoma cells), which can express different kinds of recombinant proteins. In order to reach high cell densities in these bioreactors, it is necessary to carry out an optimization of their production processes. Hence, macroscopic model equations must be developed to describe cell growth, nutrient consumption and product generation. These models will be very useful for designing the bioprocess, for developing robust controllers and for optimizing its productivity.<p>This thesis presents a new kinetic model of hybridoma cell metabolism in fed batch culture and typical illustration of a systematic methodology for mathematical modelling, parameter estimation and model-based optimization and state estimation of bioprocesses. <p>In the first part, a macroscopic model that takes into account phenomena of overflow metabolism within glycolysis and glutaminolysis is proposed to simulate hybridoma HB-58 cell cultures. The model of central carbon metabolism is reduced to a set of macroscopic reactions. The macroscopic model describes three metabolism states: respiratory metabolism, overflow metabolism and critical metabolism. The model parameters and confidence intervals are obtained via a nonlinear least squares identification. It is validated with experimental data of fed-batch hybridoma cultures and successfully predicts the dynamics of cell growth and death, substrate consumption (glutamine and glucose) and metabolites production (lactate and ammonia). Based on a sensitivity analysis of the model outputs with respect to the parameters, a model reduction is proposed. <p>In the next step, the effort is directed to the maximization of biomass productivity in fed-batch cultures of hybridoma cells based on the overflow metabolism model. Optimal feeding rate, on the one hand, for a single feed stream containing both glucose and glutamine and, on the other hand, for two separate feed streams of glucose and glutamine are determined using a Nelder-Mead simplex optimization algorithm. Two different objective functions (performance criteria) are considered for optimization; the first criterion to be maximized is the biomass productivity obtained at the end of the fed-batch culture, the second criterion to be minimized is the difference between global substrate consumption and the maximum respiratory capacity.<p>The optimal multi exponential feed rate trajectory improves the biomass productivity by 10% as compared to the optimal single exponential feed rate. Moreover, this result is validated by the one obtained with the analytical approach in which glucose and glutamine are fed to the culture so as to control the hybridoma cells at the critical metabolism state, which allows maximizing the biomass productivity. The robustness analysis of optimal feeding profiles obtained with different optimization strategies is considered, first, with respect to parameter uncertainties and, finally, with respect to model structure errors.<p>Finally, the overflow metabolism model is used to develop an extended Kalman filter for online estimation of glucose and glutamine in hybridoma cell fed-batch cultures based on the considered available measurements (biomasses (on-line), lactate and ammonia (on-line or off-line)). The observability conditions are examined, and the performances are analysed with simulations of hybridoma cell fed-batch cultures. Glutamine estimation sensitivity is enforced by minimizing a cost function combining a usual least-squares criterion with a state estimation sensitivity criterion. <p> / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished

L'estimation de la pose (position et l'attitude) en temps réel est une fonction clé pour les véhicules autonomes routiers. Cette thèse vise à étudier des systèmes de localisation pour ces véhicules en utilisant des capteurs automobiles à faible coût. Trois types de capteurs sont considérés : des capteurs à l'estime qui existent déjà dans les automobiles modernes, des récepteurs GNSS mono-fréquence avec antenne patch et une caméra de détection de la voie regardant vers l’avant. Les cartes très précises sont également des composants clés pour la navigation des véhicules autonomes. Dans ce travail, une carte de marquage de voies avec une précision de l’ordre du décimètre est considérée. Le problème de la localisation est étudié dans un repère de travail local Est-Nord-Haut. En effet, les sorties du système de localisation sont utilisées en temps réel comme entrées dans un planificateur de trajectoire et un contrôleur de mouvement pour faire en sorte qu’un véhicule soit capable d'évoluer au volant de façon autonome à faible vitesse avec personne à bord. Ceci permet de développer des applications de voiturier autonome aussi appelées « valet de parking ». L'utilisation d'une caméra de détection de voie rend possible l’exploitation des informations de marquage de voie stockées dans une carte géoréférencée. Un module de détection de marquage détecte la voie hôte du véhicule et fournit la distance latérale entre le marquage de voie détecté et le véhicule. La caméra est également capable d'identifier le type des marquages détectés au sol (par exemple, de type continu ou pointillé). Comme la caméra donne des mesures relatives, une étape importante consiste à relier les mesures à l'état du véhicule. Un modèle d'observation raffiné de la caméra est proposé. Il exprime les mesures métriques de la caméra en fonction du vecteur d'état du véhicule et des paramètres des marquages au sol détectés. Cependant, l'utilisation seule d'une caméra a des limites. Par exemple, les marquages des voies peuvent être absents dans certaines parties de la zone de navigation et la caméra ne parvient pas toujours à détecter les marquages au sol, en particulier, dans les zones d’intersection. Un récepteur GNSS, qui est obligatoire pour le démarrage à froid, peut également être utilisé en continu dans le système de localisation multi-capteur du fait qu’il permet de compenser la dérive de l’estime. Les erreurs de positionnement GNSS ne peuvent pas être modélisées simplement comme des bruits blancs, en particulier avec des récepteurs mono-fréquence à faible coût travaillant de manière autonome, en raison des perturbations atmosphériques sur les signaux des satellites et les erreurs d’orbites. Un récepteur GNSS peut également être affecté par de fortes perturbations locales qui sont principalement dues aux multi-trajets. Cette thèse étudie des modèles formeurs de biais d’erreur GNSS qui sont utilisés dans le solveur de localisation en augmentant le vecteur d'état. Une variation brutale due à multi-trajet est considérée comme une valeur aberrante qui doit être rejetée par le filtre. Selon le flux d'informations entre le récepteur GNSS et les autres composants du système de localisation, les architectures de fusion de données sont communément appelées « couplage lâche » (positions et vitesses GNSS) ou « couplage serré » (pseudo-distance et Doppler sur les satellites en vue). Cette thèse étudie les deux approches. En particulier, une approche invariante selon la route est proposée pour gérer une modélisation raffinée de l'erreur GNSS dans l'approche par couplage lâche puisque la caméra ne peut améliorer la performance de localisation que dans la direction latérale de la route. / Estimating the pose (position and attitude) in real-time is a key function for road autonomous vehicles. This thesis aims at studying vehicle localization performance using low cost automotive sensors. Three kinds of sensors are considered : dead reckoning (DR) sensors that already exist in modern vehicles, mono-frequency GNSS (Global navigation satellite system) receivers with patch antennas and a frontlooking lane detection camera. Highly accurate maps enhanced with road features are also key components for autonomous vehicle navigation. In this work, a lane marking map with decimeter-level accuracy is considered. The localization problem is studied in a local East-North-Up (ENU) working frame. Indeed, the localization outputs are used in real-time as inputs to a path planner and a motion generator to make a valet vehicle able to drive autonomously at low speed with nobody on-board the car. The use of a lane detection camera makes possible to exploit lane marking information stored in the georeferenced map. A lane marking detection module detects the vehicle’s host lane and provides the lateral distance between the detected lane marking and the vehicle. The camera is also able to identify the type of the detected lane markings (e.g., solid or dashed). Since the camera gives relative measurements, the important step is to link the measures with the vehicle’s state. A refined camera observation model is proposed. It expresses the camera metric measurements as a function of the vehicle’s state vector and the parameters of the detected lane markings. However, the use of a camera alone has some limitations. For example, lane markings can be missing in some parts of the navigation area and the camera sometimes fails to detect the lane markings in particular at cross-roads. GNSS, which is mandatory for cold start initialization, can be used also continuously in the multi-sensor localization system as done often when GNSS compensates for the DR drift. GNSS positioning errors can’t be modeled as white noises in particular with low cost mono-frequency receivers working in a standalone way, due to the unknown delays when the satellites signals cross the atmosphere and real-time satellites orbits errors. GNSS can also be affected by strong biases which are mainly due to multipath effect. This thesis studies GNSS biases shaping models that are used in the localization solver by augmenting the state vector. An abrupt bias due to multipath is seen as an outlier that has to be rejected by the filter. Depending on the information flows between the GNSS receiver and the other components of the localization system, data-fusion architectures are commonly referred to as loosely coupled (GNSS fixes and velocities) and tightly coupled (raw pseudoranges and Dopplers for the satellites in view). This thesis investigates both approaches. In particular, a road-invariant approach is proposed to handle a refined modeling of the GNSS error in the loosely coupled approach since the camera can only improve the localization performance in the lateral direction of the road. Finally, this research discusses some map-matching issues for instance when the uncertainty domain of the vehicle state becomes large if the camera is blind. It is challenging in this case to distinguish between different lanes when the camera retrieves lane marking measurements.As many outdoor experiments have been carried out with equipped vehicles, every problem addressed in this thesis is evaluated with real data. The different studied approaches that perform the data fusion of DR, GNSS, camera and lane marking map are compared and several conclusions are drawn on the fusion architecture choice.

<p>Doktorskom disertacijom je dokazano da postojeće metode detekcije i identifikacije loših podataka nisu primenjive na distributivne mreže usled njihovih specifičnosti u stepenu redundanse merenja i broja pseudo merenja. Dodatno, razvijeni su algoritmi detekcije loših oblasti primenom dekuplovanog Hi-kvadrat testa, identifikacije loših merenja primenom novo definisanih izbeljenih reziduala, estimacije fazne konektivnosti primenom uslovnih ograničenja u estimatoru stanja, i korekcije pseudo merenja primenom informacija sa pametnih brojila. Navedeni algoritmi su specijalizovani za distributivne mreže i verifikovani primenom na dva test sistema.</p> / <p>The doctoral dissertation has demonstrated that conventional bad data detection and<br />identification methods cannot be efficiently applied in distribution networks, due to<br />their characteristics such as low measurement redundancy, number of pseudo<br />measurements and level of measurements correlation. In addition, the doctoral<br />dissertation described newly developed algorithms for bad area detection based on<br />decoupled Chi-squares test, bad data identification using newly defined whitened<br />residuals, estimation of phase connectivity by extension of state estimation with<br />conditional constraints and correction of pseudo measurements using AMI data. The<br />mentioned algorithms are specialized for distribution networks and verified through<br />simulation on two test systems.</p>

No description available.

Dans le futur, les réseaux de distribution deviendront intelligents et actifs et seront utilisés au plus près de leurs limites car l'avènement de la dérégulation avec l'introduction massive de la production décentralisée induira une optimisation des infrastructures de l'énergie pour des besoins environnementaux d'une part, mais aussi par l'introduction de la concurrence dans un secteur autrefois monopolistique. Si on veut utiliser pleinement le potentiel des productions décentralisées dans les réseaux de distribution, on devra rendre ces réseaux observables au moyen de mesures qui seront intégrées dans les systèmes de supervision et de contrôle car sans observabilité et reconstruction d’état du réseau, le contrôle du réseau est impossible. Pour un grand réseau de distribution (quelques milliers de nœuds), avec un taux d’insertion important de productions décentralisées, l’observabilité en temps-réel devient très difficile. Dans la littérature scientifique, on ne trouve pas de travaux prenant en compte les réseaux intelligents et reconfigurables de demain avec production décentralisée massive. L'enjeu du projet sera donc de développer des algorithmes de reconstruction d'état prenant en compte les spécificités des réseaux du futur, à les valider et à les intégrer dans les outils de gestion des réseaux du distributeur. La reconstruction d'état permettra d'aider à l'automatisation et donc à l'introduction d'intelligence dans les réseaux de distribution du futur ainsi qu'à l'insertion massive de productions décentralisées.________________________________________ / In the future, distribution networks will become intelligent and active and will be operated as close as possible to their limits, with the advent of deregulation and with the introduction of mass production that will lead to a decentralized infrastructure, this will guide to an optimization of energy for environmental issues in one side, but also to the introduction of competition in a sector once monopolistic. If the potential of distributed generation in the distribution networks is to be fully exploited, we will make these networks observable through measures which will be integrated into supervision systems and control, because without observability and reconstruction of network status, control network is impossible. For a large distribution network (a few thousand nodes), with an important quantity of distributed generation, real-time observability becomes very difficult. In the scientific literature, there is no work taking into consideration intelligent networks and reconfigurable of the future with decentralized mass power generation. The challenge of the project will be to develop algorithms of reconstruction of the state of the network, taking into account the characteristics of the networks of the future, to validate and integrate them into management tools distribution networks. The reconstruction of the state will assist in the automation and hence the introduction of intelligence in the distribution networks of the future and the insertion massive distributed generation.

With the increased performance and reduced cost of cameras, the robotics community has taken great interest in estimation and control algorithms that fuse camera data with other sensor data.In response to this interest, this dissertation investigates the algorithms needed for robust guidance, navigation, and control of fixed-wing and multirotor aircraft applied to target estimation and circumnavigation.This work begins with the development of a method to estimate target position relative to static landmarks, deriving and using a state-of-the-art EKF that estimates static landmarks in its state.Following this estimator, improvements are made to a nonlinear observer solving part of the SLAM problem.These improvements include a moving origin process to keep the coordinate origin within the camera field of view and a sliding window iteration algorithm to drastically improve convergence speed of the observer.Next, observers to directly estimate relative target position are created with a circumnavigation guidance law for a multirotor aircraft.Taking a look at fixed-wing aircraft, a state-dependent LQR controller with inputs based on vector fields is developed, in addition to an EKF derived from error state and Lie group theory to estimate aircraft state and inertial wind velocity.The robustness of this controller/estimator combination is demonstrated through Monte Carlo simulations.Next, the accuracy, robustness, and consistency of a state-of-the-art EKF are improved for multirotors by augmenting the filter with a drag coefficient, partial updates, and keyframe resets.Monte Carlo simulations demonstrate the improved accuracy and consistency of the augmented filter.Lastly, a visual-inertial EKF using image coordinates is derived, as well as an offline calibration tool to estimate the transforms needed for accurate, visual-inertial estimation algorithms.The imaged-based EKF and calibrator are also shown to be robust under various conditions through numerical simulation.

Les nouvelles générations de radars sont confrontées à des cibles de plus en plus menaçantes. Ces radars doivent effectuer plusieurs tâches en parallèle, dont la veille et la poursuite. Pour cela, ils peuvent être équipés de panneaux fixes, pour éviter les contraintes liées à la rotation de l'antenne. Le pistage du radar doit donc être renouvelé pour répondre à la double difficulté posée par le pistage des cibles très manoeuvrantes et la gestion des ressources. Dans ce contexte, cette thèse étudie de nouvelles méthodes de pistage pour les cibles hyper-manoeuvrantes. Un nouveau modèle de cible, en coordonnées intrinsèques, est proposé. Ce modèle est exprimé directement dans le repère de la cible, afin de décrire au mieux des manoeuvres fortes avec des accélérations normales bien supérieures à la gravité terrestre. Un algorithme de filtrage utilisant la formulation intrinsèque du modèle est développé. Cet algorithme ayant la même structure qu'une filtre de Kalman étendu, il a été testé sur de vraies données. La comparaison avec d'autres algorithmes de filtrage a montré de réelles améliorations sur un ensemble important de trajectoires. Une nouvelle méthode d'estimation, reposant sur la formulation en termes de moindres carrés de l'approche de lissage, et permettant de tenir compte de sauts dans la trajectoire est également proposée, et les bénéfices sur des méthodes plus classiques de sauts entre modèles sont montrés. Indépendamment, le problème de cadence adaptative est également traité. Un algorithme très général permettant d'optimiser la cadence de mesure pour ménager le budget temps du radar pour la surveillance est présenté. / The new generation of radars is facing increasingly threatening targets. These radars are asked to perform several tasks in parallel, including surveillance and tracking. To this aim, they can be equipped with staring antennas, so they overcome the constraints induced by the rotation of the antenna. The tracking function of the radar has thus to be upgraded to respond to the double issue of tracking highly manoeuvring targets and managing the resources to balance time between tasks. In this context, this thesis investigates new means of tracking highly manoeuvring targets. A new target model based on intrinsic coordinates to perform target tracking is proposed. This new target model is expressed in the frame of the target itself, and uses the Frenet-Serret frame, which is well suited to the description of highly dynamic manoeuvres involving normal accelerations that are much larger than earth gravity. A filtering algorithm using the special intrinsic formulation of the target model is developed. This filtering algorithm is very similar in terms of implementation to an Extended Kalman filter, and was implemented using real data. The comparison with standard target models and filtering algorithms show improvements over simple models and algorithms on a large set of trajectories. A new estimation method, relying on the least squares formulation of the smoothing approach, and taking into account kinematic jumps in the trajectory is also developed. This method also shows improvements over a set of common algorithms based on standard manoeuvre detection. And independently, we investigate the issue of update rate adaptation for radar measurements. A very general update rate adaptation algorithm is derived to optimise the time of revisit of each target, allowing to preserve the radar time budget for other tasks simultaneously performed, such as surveillance.

<div>The electric power system of a deep space vehicle is mission-critical, and needs to operate autonomously because of high latency in communicating with ground-based mission control. Key tasks to be automated include managing loads under various physical constraints, continuously monitoring the system state to detect and locate faults, and efficiently responding to those faults. </div><div><br></div><div>This work focuses on three aspects for achieving autonomous, fault-tolerant operation in the dc power system of a spacecraft. First, a sequential procedure is proposed to estimate the node voltages and branch currents in the power system from erroneous sensor measurements. An optimal design for the sensor network is also put forth to enable reliable sensor fault detection and identification. Secondly, a machine-learning based approach that utilizes power-spectrum based features of the current signal is suggested to identify component faults in power electronic converters in the system. Finally, an optimization algorithm is set</div><div>forth that decides how to operate the power system under both normal and faulted conditions. Operational decisions include shedding loads, switching lines, and controlling battery charging. Results of case studies considering various faults in the system are presented.</div>