Nonlinear Output Stabilization of Uncertain Systems

Recently, a great deal of attention has been paid to stabilization of nonlinear or uncertain systems. Most of the proposed methods require access to all of the states for implementation. However, this requirement cannot be satisfied in general. This paper offers an alternative stabilization scheme whereby the output of the system is fed back. This method is contrasted with a full-state feedback strategy using observed states. The design is illustrated utilizing a single flexible link manipulator where compliance of the link and other non-idealities are treated as a lumped nonlinearity/uncertainty.