Robust nonlinear control of feedforward systems with unmodeled dynamics

From MaRDI portal





Nonlinear continuous time systems in feedforward form are considered with unmodeled dynamics appearing at the system input. More precisely, the dynamics of the nominal system is described by the equation \(\dot x=f( x,v,d) \), the unmodeled dynamics is represented by \(\dot \xi=q( \xi ,u,d) \) and \( v=p( \xi ,u,d) \), where \(d\) is a disturbance, \(u\) is the input of the system and \(v\) is the input of the nominal system. For this class of systems a nested saturation control law \(u=-\varphi _{1}( k_{1}x_{1}+\varphi _{2}( k_{2}x_{2}+\ldots \varphi _{n}( k_{n}x_{n}) \ldots)) \), where \(\varphi _{i}( .) \) are saturation functions and \(k_{i}\) are controller gains, is defined and a procedure is provided for the selection of \(k_{i}\) and saturation levels to guarantee global asymptotic stability, in the absence of disturbance, and finite disturbance to state gain, if the disturbance is bounded. The design does not require detailed knowledge of the unmodeled dynamics, and no restrictions on their zero dynamics and relative degree are needed. The authors show that the achieved robustness property is due to the law gain design, which is sufficient for the stabilization of systems in feedforward form. Without this feedforward structure, global stabilization might not be possible, using the same design tools.




Cited in
(23)








This page was built for publication: Robust nonlinear control of feedforward systems with unmodeled dynamics

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5926175)