Recent Projects
IMPROVED Project: Intelligent Maintenance & Prognostics via real-time Estimations and control Decisions (2015-2020)
New methodologies for modeling and
control of mechanical deterioration. In particular, control of the
Remaining Useful Lifetime (RUL) on friction-based actuators. It
concerns a new research topic at GIPSA-lab which combines Automatic
control theory and Reliability engineering. In collaboration with:
Christophe Berenguer (Prof at Grenoble-INP) and Diego Rodriguez (Ph.D.
student at GIPSA-lab). This project has been financed by the program
PEPS CNRS INS2I 2017.
Ebike4.0: Physiology-aware control for Ebikes (2015-2020).
This project concerns a scientific effort for designing a new generation of electronic power assisted bicycles. In particular we focus on the Physiology-aware control of pedelecs, where the bicycle-rider system is modeled as a Linear Parametric Varying (LPV) system. This kind of models takes into account the effect of different operation points, nonlinearities and parametric uncertainties. Specifically, we work in modeling the behavior of aerobic and anaerobic reactions for producing mechanical energy at the pedal level. It concerns a new research topic at GIPSA-lab which combines Automatic control theory and Physiology - Bioenergetics. In collaboration with: Nadia Rosero (Ph.D. student at GIPSA-lab), Matteo Corno (Prof. at Politecnico di Milano, Italy), Henry Leon (M.D. at the Universidad de la Sabana, Colombia) and Johan Cassirame (Ph.D. and sales manager at Matsport).
Set-membership state observer design for fault-tolerant control (2014-2019)
This project concerns the design and
the implementation of accurate Set-membership state observers for
linear and LPV systems. In particular we are interested in developing
more easily implementable algorithms than those available into the
literature. We explore the use of positive invariant sets for
obtaining explicit descriptions of the estimation-error bounds. Such
estimation-error bounds are used for implementing the set-membership
state observer in a more explicit way. The observer design is mostly
based on ellipsoidal positive invariant sets obtained by using the
Bounded-real lemma which is well known into the Robust control theory.
In collaboration with: Nassim Loukkas (Ph.D. student) and Nacim Meslem
(GIPSA-lab). Set-invariance approaches for Fault-Tolerant Control
concerns a join research work in collaboration with Maria Seron and
Jose De Dona (University of Newcastle, Australia) and, Ernesto Kofman
and Hernan Haimovich (Universidad del Rosario, CIFASIS - CONICET,
Argentina).
Event-based control for robotic systems : A set-invariant based approach (2013-2017)
In this project it is proposed a new
event-switched control method for controlling discrete-time linear
systems subject to bounded disturbances and general robotic systems.
The main advantage of the proposed method is that the nominal
performance of the controlled system with periodic control updates is
kept in a framework that do not require to periodically update the
control law. The feedback control loop can be opened as long a
state-dependent event condition is satisfied. This condition is
obtained using set theory approaches. In particular, the concept of
robustly positively invariant sets is used to calculate the nominal
performance and the event condition. In collaboration with: Bruno
Boisseau (a former Ph.D. student), Nicolas Marchand (GIPSA-lab),
Thibaut Raharijaona (Univ. Aix-Marseille) and Sylvain Durand (INSA
Strasbourg). Today a software using this technology is tested for
controlling drones in the context of a project financed by Linksium
SATT Grenoble. See Drone Interactive (Vincent Rigau):
FORSENS: Fast real-time force estimation method for rotating systems (2015-2018)
Real-time force and torque estimation
method for rotating system and pedelecs. Project financed by Linksium
SATT Grenoble and Grenoble-INP (France). A patent application has been
submitted in 2016 and granted in 2018. Today a software using this
technology is tested for solving advanced motion control problems.
Former collaborator: Ricardo Rodriguez (2015-2016).
Other Projects (Before 2013)
Switching control theory and Fault-tolerant control (2005-2013)
Fault-tolerant control for automotive
vehicles. ANR INOVE Project. Coordinator: Olivier Sename. In
collaboration with: Laboratoire MIPS, Mines ParisTech, and the
industrial partner SOBEN (France). Responsible of the WP :
Fault-detection and Fault-Tolerant Control.
Robust and Optimal control (2008-2012)
Identification and Robust Rejection of Vibrations. In collaboration with Ioan Landau (CNRS, Gipsa-lab). Industrial cooperation: Hutchinson (Vibrachoc).Robust control of an embedded system (Drones). In collaboration with Nicolas Marchand (CNRS, Gipsa-lab). ANR EVA Project.
Coordination of distributed MPC for Hydro-Power Valley control. In collaboration with Gildas Besancon (Gipsa-lab). Industrial cooperation: Serge Maurin (EDF - R&D de Chatou).
Development of an exoskeleton control system. In collaboration with José Ramírez Scarpetta. Universidad del Valle, Grupo GICI Cali Colombia. Supported by COLCIENCIAS.
Hydro-Power Valley optimal control using MPC. In collaboration with José Ramírez Scarpetta. Universidad del Valle, Grupo GICI Cali Colombia. Supported by COLCIENCIAS.
European BLAZE project (2005-2008)
The BLAZE project aims at defining an
overall architecture and development of components for the Blu-ray HD
recorder. See MEDEA+. See 2A201_BLAZE. Industrial cooperation: Jos
Huisken, Philips Research - Silicon Hive (Holland). Andrew Dumbill and
Peter Pohl, DaTARIUS (Austria).
The BLAZE project has received the Jean-Pierre
Noblanc award for excellence 2008.French ARCOS Program (2001-2004)
National Program on security roads and
vehicles. PREDIT projet. See also Inter-distances . See:
www.arcos2004.com. Scientific adviser: Jean-Mark Blosseville, LIVIC
laboratory of the INRETS/LCPC (France).
Industrial adviser: Xavier Claeys, Renault (France).