Thesis of Achille Melingui

Modélisation et commande d'une classe de mobiles omnidirectionnels - robots manipulateurs continus: Cas du Robotino XT

This dissertation addresses the modeling and the control of a class of mobile and continuum manipulator robots. As a case study, we focused on the system of Robotino XT, developed in the framework of industrial cooperation. It consists of a bionic continuum manipulator mounted on an omnidirectional mobile platform named Robotino. Thus, we first propose a hybrid adaptive controller, based on Type-2 Fuzzy Logic (type-2 FL) and Artificial Potential Field (APF) to navigate accurately the mobile platform. Then, an adaptive neural network controller is proposed for the manipulator part. The latter includes two sub-controllers. The first one, associated to the manipulator kinematics, is based on Distal Supervised Learning (DSL) scheme, dealing with stationary behaviors of the manipulator; while the second, associated to the manipulator kinetics, is based on adaptive control, aiming to compensate for the non stationary behaviors. Finally, the two controllers are coordinated by a neural network system to control the Robotino XT. The performance of each controller is assessed through a set of real-time experiments.

Jury

Véronique PERDEREAU Professeur, Université Pierre et Marie Curie (France) Rapporteur Abdelaziz BENALLEGUE Professeur, Université de Versailles S. Q. Y. (France) Rapporteur Belkacem Ould Bouamama Professeur, Université Lille1, (France) Examinateur Christian Duriez Directeur de recherches, Inria Nord-Europe (France) Examinateur Boubaker Daachi Maître de conférences, Université Paris Est (France) Examinateur EMMANUEL TONYE Professeur, Université Yaoundé1, (France) Examinateur Rochdi Merzouki Professeur, Université Lille1, (France) Directeur de thèse Jean Bosco MBEDE Chargé de Cours, Université de Yaoundé1, (Cameroun) Co-directeur de thèse

Thesis of the team defended on 09/09/2014