Adaptive output-feedback control of a class of nonlinear systems

For a class of single-input-single-output nonlinear systems with unknown constant parameters, the authors construct a novel systematic procedure for adaptive nonlinear control design, which requires only output, rather than full-state, measurement and which yields global boundedness and tracking properties without imposing any type of growth constraints on the nonlinearities. The proposed procedure is applicable to nonlinear systems which can be expressed in the output-feedback canonical form. The authors give a coordinate-free characterization of this class of systems, and show that a single-link robotic manipulator with an elastically coupled DC-motor actuator belongs to this class, and can thus be adaptively controlled via the design procedure using only position measurement