Aerial manipulation, Nonlinear control, State estimation, Multi-agent systems, Cooperative Control, Consensus algorithms, Even-based control.
Multi-MAV Formation Control for Interactive Operations
My research activities in the field of Multi-agent systems for Interactive Operations lies in the proposition and validation of control and navigation strategies for the autonomous grasping and transportation based on a fleet of mini aerial vehicles (MAV's) equipped with robotic interactive means. A preliminary experimental stage (proof-of-concept benchmark) will validate the proposed guidance/navigation/control (GNC) strategies.
My activities within the project include:
Modeling and control of the aerial vehicles
Formation control of the multi-mav system
Proposition of novel interactive robotic means
Modeling and Control of VTOL Vehicles with Rigid Manipulators
My Ph.D. thesis, supervised by Nicolas MARCHAND, Fermi GUERRERO-CASTELLANOS and Sylvain DURAND, has been carried out at the GIPSA-lab in Grenoble, France. My research activity was focused on the modeling, control and implementation of an aerial vehicle (multirotor) carrying a manipulator arm.
The Ph.D. thesis proposes a simple solution for the stabilization of a mini-quadcopter carrying a 3DoF (Degrees of Freedom) manipulator robot in order to enhance its achievable workspace and application profile. Since the motion of the arm induces torques which degrade the stability of the system, the present dissertation consideres the stabilization of both subsystems: the quadcopter and the robotic arm. The mathematical model of the system is based on quaternions. Likewise, an attitude control law consisting of a bounded quaternion-based feedback stabilizes the quadcopter to a desired attitude while the arm is evolving. The next stage is the translational dynamics which is simplified for control (nonlinear) design purposes. The aforementioned controllers are based on saturation functions whose stability is explicitly proved in the Lyapunov sense. Finally, experimental results and a statistical study validate the proposed control strategy.
The dissertation is available here: DISSERTATION
A video of the experimental platform is available here: