Title :
Controller Design for Systems with Unknown Nonlinear Dynamics
Author :
Youcef-Toumi, Kamal ; Ito, Osamu
Author_Institution :
Assistant Professor, Laboratory for Manufacturing and Productivity, Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139 U.S.A.
Abstract :
The control of systems with unknown dynamics is an important issue, particularly in applications where large parameter variations and unexpected disturbances are possible. Several methods have been developed for such cases. These include adaptive control, variable structure control and learning control. This paper first defines this control problem and then proposes a new method - Model Reference Control Using Time Delay (MRC/TD) for nonlinear plants with unknown dynamics. This method will not requireparameter identification, unnecessary discontinuous control or repetative actions. The main feature of the MRC/TD is the direct estimation of the unknown dynamics and the unexpected disturbance. This is accomplished using time delay. This estimation is then used by the controller to cancel both the undesired dynamics and disturbance and to insert the desired dynamics into the plant. The MRC/TD control law for nonlinear plants is presented and the stability analysis and design procedure are given for SISO linear plants. Finally, the validity of MRC/TD is verified through both simulations and experiments.
Keywords :
Adaptive control; Control systems; Delay effects; Delay estimation; Error correction; Manipulator dynamics; Mechanical variables control; Nonlinear control systems; Robots; Vehicle dynamics;
Conference_Titel :
American Control Conference, 1987
Conference_Location :
Minneapolis, MN, USA