Automatic Control and MicroMechatronic Systems Department, FEMTO-ST Institute, Université Bourgogne Franche-Comté, CNRS, 24, rue Alain Savary, Besançon 25000, France
Auto-generated profile, based on publicly available information
Mounir Hammouche
Automatic Control and MicroMechatronic Systems Department, FEMTO-ST Institute, Université Bourgogne Franche-Comté, CNRS, 24, rue Alain Savary, Besançon 25000, France
Controlling multidegrees of freedom (DOF) micropositioning systems always represents a great challenge because of the high sensitivity to the environment at this scale and the cross-coupling effects present between the different axes. A robust Luenberger observer-based state feedback design using interval analysis and the regional pole assignment technique are introduced to control such systems. This robust control design keeps the same structure of the classical state-feedback with the usual Luenberger observer. However, the syntheses of the observer and the feedback controller are performed by means of interval techniques to find the set of gains that are robust against system uncertainties and that satisfy some predefined performances. For this matter, an algorithm based on set inversion via interval analysis (SIVIA) combined with interval eigenvalues computation is proposed to find these robust gains. The control approach is validated in simulation and then tested experimentally to control a multi-DOF positioning structure.
5 June 2018 in IEEE Transactions on Cognitive and Developmental Systems
Human Action recognition (HAR) is largely used in the field of Ambient Assisted Living (AAL) to create an interaction between humans and computers. In these applications, it cannot be asked to people to act non-naturally. The algorithm has to adapt and the interaction has to be as quick as possible to make this interaction fluent. To improve the existing algorithms with regards to that points, we propose a novel method based on skeleton information provided by RGB-D cameras. This approach is able to carry out early action recognition and is more robust to viewpoint variability. To reach this goal, a new descriptor called Body Directional Velocity is proposed and a real-time classification is performed. Experimental results on four benchmarks show that our method competes with various skeleton-based HAR algorithms. We also show the suitability of our method for early recognition of human actions.
The problem of robust and optimal output feedback design for interval state-space systems is addressed in this paper. Indeed, an algorithm based on set inversion via interval analysis (SIVIA) combined with interval eigenvalues computation and eigenvalues clustering techniques is proposed to seek for a set of robust gains. This recursive SIVIA-based algorithm allows to approximate with subpaving the set solutions [K] that satisfy the inclusion of the eigenvalues of the closed-loop system in a desired region in the complex plane. Moreover, the LQ tracker design is employed to find from the set solutions [K] the optimal solution that minimizes the inputs/outputs energy and ensures the best behaviors of the closed-loop system. Finally, the effectiveness of the algorithm is illustrated by a real experimentation on a piezoelectric tube actuator.
International audienceIn this paper we propose a novel human action recognition method, robust to viewpoint variation, which combines skeleton-and depth-based action recognition approaches. For this matter, we first build several base classifiers, to independently predict the action performed by a subject. Then, two efficient combination strategies , that take into account skeleton accuracy and human body orientation, are proposed. The first is based on fuzzy switcher where the second uses a combination between fuzzy switcher and aggregation. Moreover, we introduce a new algorithm for the estimation of human body orientation. To perform the test we have created a new Multiview 3D Action public dataset with three viewpoint angles (30°,0°,-30°). The experimental results show that an efficient combination strategy of base classifiers improves the accuracy and the computational efficiency for human action recognition