PDE Boundary Control for Euler-Bernoulli Beam Using a Two Stage Perturbation Observer

A novel perturbation observer-based PDE boundary control law for beam bending is derived based on a combination of perturbation observers and polynomial trajectory planning. The perturbation observer consists of two components. The first stage employs the "particular" solution of the original dynamics with disturbances while its boundary conditions are set to zero. In contrast, the dynamics of the "homogeneous component" are independent of the beam dynamics, but its boundary conditions are identical to those of the beam. A tracking boundary control law, based on trajectory planning, is designed for the homogeneous component, and the same control signal is also applied to the beam. The stability of the adaptive perturbation-observer is proven by Lyapunov stability in the spatial L-2 sense, while stability conditions are derived for a finite dimensional ODE analogue of the infinite dimensional closed loop PDE system. This paper also reports on one of the first experimental demonstrations of a controller designed entirely using a PDE boundary control formulation.