Identification, estimation and control for linear systems using measurements of higher order derivatives

Zilong Shao 1 Gang Zheng 1, 2 Denis Efimov 1 Wilfrid Perruquetti 1
1 NON-A - Non-Asymptotic estimation for online systems
Inria Lille - Nord Europe, CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189
2 DEFROST - Deformable Robots Simulation Team
Inria Lille - Nord Europe, CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189
Abstract : The problem of output control for linear uncertain systems with external perturbations is studied. Instead of the commonly used assuption that the state variable is mesurable, here it is assumed that the output available for measurements is the higher order derivative of the state only (for example, the acceleration for a second order plant), which is also corrupted by noise. Then via series of integration an identification algorithm is proposed for identification of values of all parameters and unknown initial conditions for the state vector. Finally, two control algorithms are developed, adap-tive and robust, providing boundedness of trajectories for the system. Efficiency of the obtained solutions is demonstrated by numerical experiments. 1 Introduction The problem of output control for uncertain linear systems with external perturbations is one of the most important problems in the control theory. A model-based controller design requires an identification process of the uncertain system , for which full state variables need to be either measured or estimated. Usually observers for such systems are designed under assumption that only the outputs are available for measurement but not their derivatives [1]. However, there are cases where measurement of high-order derivatives of the output state can be more convenient than that of output state itself, for example, it is much more easier for an individual to attach an accelerometer to the end-effector of a robot manipulator than to mount a position decoder to the motor rotor inside the robot body. This paper addresses the mentioned problem under assumption that only the higher order derivative of the state is measurable. In the present work, we are mainly motivated by identification , estimation and control in a robot manipulator application , when sensors are installed in the robot motors at the basements of joints, but due to flexibility of joints, the end-point position of the joint does not always coincide with the value given by motor sensor under rigid geometry of the robot. In order to improve accuracy of estimation and control in such a case, an accelerometer can be installed at the
Type de document :
Article dans une revue
Journal of Dynamic Systems, Measurement, and Control, American Society of Mechanical Engineers, 2017, 139 (12), pp.1-6. 〈10.1115/1.4037007〉
Liste complète des métadonnées

Littérature citée [6 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01649537
Contributeur : Gang Zheng <>
Soumis le : lundi 27 novembre 2017 - 17:01:58
Dernière modification le : mercredi 4 juillet 2018 - 17:36:13

Fichier

Final_Manuscript_File_ASME.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Zilong Shao, Gang Zheng, Denis Efimov, Wilfrid Perruquetti. Identification, estimation and control for linear systems using measurements of higher order derivatives. Journal of Dynamic Systems, Measurement, and Control, American Society of Mechanical Engineers, 2017, 139 (12), pp.1-6. 〈10.1115/1.4037007〉. 〈hal-01649537〉

Partager

Métriques

Consultations de la notice

159

Téléchargements de fichiers

30