Research
Feedback Linearizable Discretizations for Mechanical Systems with Retraction Maps
Mechanical systems, especially nonlinear systems have the dynamics (or spray) evolving on a manifold which has a particular mechanical structure. Feedback linearization is a technique to design feedback controllers for nonlinear systems, in such a way that the constructed feedback control allows for a locally linear system at every time instant. So, the question is now about how one can construct a feedback linearizable discretization which preserves the mechanical structure of the system.
Nonlinear Infinite-Dimensional Modeling and Control of a Two DoF Flexible Wing
A flexible wing can be modeled in many different ways, and for different applications. The aeroelastic effects create many interesting phenomena, especially flutter, which must be considered in control design. If one assumes a high aspect-ratio wing, bending and twisting are the two relevant degrees of freedom which can be considered in the derivation of the dynamic equations in the form of partial differential equations.
Design and Control of Flapping Wing Mechanism
Bio-mimickry in aerospace engineering has been the main motivation in the industry due to the natural efficiency found in avians, achieved over years of evolution. Our flapping wing mechanism is a design utilizing mechanically coupled feathering-flapping 2-DoF mechanism. The design is made such that the phase difference between flapping and feathering is 90 degrees and the aerodynamics is mathematically modeled to develop a controller.
Safety-Critical Control for Unmanned Aerial Vehicles
Unmanned aerial vehicles (UAVs) are a booming technology in the industry today. A highly ambitious application would be to design controllers for usage of a UAV in a Mars-like environment. Designing controllers for a system like a quadcopter, involves many state and control constraints. Popular techniques include Model Predictive Control (MPC) and Model Reference Adaptive Control (MRAC). One could try to implement MRAC with state and control constraints. For real-world applications, some safety bounds must be considered, which can be done using barrier Lyapunov functions (BLFs). Thus, we are interested in safety-critical control applied to a UAV in Mars-like environment, and validate this in Gazebo-Simulink simulations.