Abstract:

Analytic tools based on differential manifolds provide an elegant and general treatment of robot kinematics, bringing significant insight into robotic nonlinear control theory which includes much of differential geometry. The nonlinear model predictive control exhibits advantages of high tracking performance and constraint handling, by solving a constrained optimal problem in a finite receding horizon. In this seminar, the speaker will present the recent research progress about embedding manifold structure into model predictive control for robot trajectory tracking.