Vous êtes ici : GIPSA-lab > Animation > Séminaires

Séminaire du département Automatique du 06/06/2013 à 14h00


Session spéciale doctorants de juin 2013

Intervenant : Ahmed, Sandoval-Moreno, Hamaz, Bouchair, Leon, Rubio-scola,

Lieu : Salle des séminaires, Département Auto., ENSE3, Bat 3, 2è étage


Résumé :

M. Ahmed: “Control Optimization of relaxation-Cycle electricity generation systems

The presentation deals principally with the grid connection problem of a kitebased system, named the “Kite Generator System (KGS).” It presents a control scheme of a closed-orbit KGS, which is a wind power system with a relaxation cycle. Such a system consists of a kite with its orientation mechanism and a power transformation system that connects the previous part to the electric grid. Starting from a given closed orbit, the optimal tether’s length rate variation (the kite’s tether radial velocity) and the optimal orbit’s period are found. The trajectory-tracking problem is solved by applying a predictive control strategy; the orientation of the kite controlled through the roll angle, while the kite’s tether radial velocity is controlled via the electric machine rotation velocity. The power transformation system transforms the mechanical energy generated by the kite into electrical energy that can be transferred to the grid. A Matlab/simulink model of the KGS is employed to observe its behavior, and to insure the control of its mechanical and electrical variables. In order to improve the KGS’s efficiency in case of slow changes of wind speed, a maximum power point tracking (MPPT) algorithm is proposed.

 J. Sandoval-Moreno: « Observer-based Maximum Power point Tracking Control Strategy for Sensorless wind Turbine Generators »

 In this talk, it will be presented a Maximum Power Point Tracking (MPPT) control strategy for a wind turbine generation system, with an important characteristic: there are only measurements from the electrical generator variables. In other words, there is no wind speed measurement. The algorithm, in a first stage, uses a Kalman-like observer for estimating the wind turbine power coefficient characteristic (related with the wind speed), that is used for the power generator torque control, as well as for the optimization algorithm. The second stage uses the results from the observer and updates an uncertain wind turbine’s power coefficient polynomial (typically obtained from the manufacturer, for example) by means of a Recursive Least-Square Algorithm. From this updated polynomial, it is possible to obtain the optimal electrical generator speed reference that maximizes the power produced by the turbine. As part of the talk, some results and remarks for future works will be presented. The talk also will be used to show how these results can be applied to alternative distributed power generation systems, that considerate wind turbines as part of the generation units, which can also include fuel cells, photovoltaic, geothermal, hydroelectrical, among others.  

 T. Hamaz: « Détection d’hétérogénéités dans une pile à combustible de type PEM pour le diagnostic»

 Les piles à combustible (PAC) permettent de transformer directement l'énergie chimique de l'hydrogène en énergie électrique et thermique. Les recherches réalisées dans le domaine du diagnostic ont pour objectif d'améliorer leur durée de vie. Nous explorons une approche par mesure du champ électromagnétique généré par la pile, cette méthode a l'avantage de proposer un diagnostic local. La mesure du champ magnétique est l'un des moyens utilisé pour identifier la distribution de la densité de courant à l'intérieur d’un stack. Les données sont constituées des valeurs instantanées des capteurs qui ne détectent que les hétérogénéités des densités de courant. Chaque jeu de mesures est représenté graphiquement et la forme est représentative de ces hétérogénéités qui peuvent être définies en 3D. Le problème de diagnostic posé est un problème de Reconnaissance des Formes. Il consiste à décider à quelle classe appartient une nouvelle observation parmi les hétérogénéités prédéfinies. Une méthode basée sur la génération de résidus vectoriels a été développée et permet l’extraction des zones d’hétérogénéités. Une autre méthode spécifique basée sur la génération des pseudos inverses des bases prédéfinies est en cours de développement, elle permet de générer des résidus scalaires afin de faciliter l’étape de décision. L’étape finale est de relier ces hétérogénéités à un mode de fonctionnement de la pile.

 N. Bouchair: “Fault diagnosis and help on operation using event list: Application on the Hadron Collider Industrial Control System”

A decision support system for operators monitoring complex systems will be exposed in this talk. It consists in a fault isolation method based on
pattern matching using binary information, such as event lists or alarm lists. A training set composed of faults is used to create fault templates.
Event lists generated by unknown faults are classified by comparing them with the fault templates using different type of similarity functions. First
the problematic will be introduced, then the method will be exposed and finally the performances of the method applied on a CERN Large Hadron
Collider process will be shown.

 L. Leon Ojeda: “Adaptive Kalman Filtering for Multi-step ahead Traffic flow prediction”

Given the importance of continuous traffic flow forecasting in most of Intelligent Transportation Systems (ITS) applications, where every new traffic data become available in every few minutes or seconds, the main objective of this study is to perform a multi-step ahead traffic flow forecasting that can meet a trade-off between accuracy, low computational load, and limited memory capacity. To this aim, based on adaptive Kalman filtering theory, two forecasting approaches are proposed. We suggest solving a multi-step ahead prediction problem as a filtering one by considering pseudo-observations coming from the averaged historical flow or the output of other predictors in the literature. For taking into account the stochastic modeling of the process and the current measurements we resort to an adaptive scheme. The proposed forecasting methods are evaluated by using measurements of the Grenoble south ring.

 I. Rubio Scola: «Online Observability for State Affine Systems with Output Injection and Observer Design»

Observability being in general subject to the applied input for a nonlinear system, the aim of the present work is to propose an input selection strategy for a special class of systems, so as to make them observable. The considered systems are those admitting a state-affine structure with output injection. For such systems, an online algorithm is proposed to compute an appropriate input in real time. It guarantees observability by ensuring a lower bound on the related Gramian, and minimizes at the same time the input variations with respect to some reference value required for the system operation. This computation is updated at each time with the new output measurements becoming available. The proposed methodology is illustrated on a piping system example, for which an exponential observer is finally obtained.

GIPSA-lab, 11 rue des Mathématiques, Grenoble Campus BP46, F-38402 SAINT MARTIN D'HERES CEDEX - 33 (0)4 76 82 71 31