Abstract
Control of a nonlinear system is addressed by considering its family of linearizations. Given that a family of linear state-feedback controllers has been designed to meet some (linear) control objective, it is shown how these controllers can be pieced together into an overall, nonlinear, state-feedback controller. The resulting closed-loop system has the property that its linearizations are equal to the desired linearly controlled systems. A similar approach can be used to design nonlinear state estimators. The approach is illustrated by applying it to the control of a planar two-link manipulator.

This publication has 4 references indexed in Scilit: