Feedback linearization using approximate models

摘要

The paper presents results developed to control sampled data nonlinear systems. It is shown that there are fundamental problems when attempting to use the standard feedback linearization procedure for sampled data systems. The solution needs some structural analysis and the introduction of state dependent parametrization. In case of unknown nonlinearities approximate models can be used in the form of spline surfaces, fuzzy logic or neural nets. Any of these approximation techniques are applicable to expand the control algorithm for adaptive systems.