LPT-2004-33 [BibTeX]
Jürgen Hahn, Martin Mönnigmann, Wolfgang Marquardt:
A method for robustness analysis of controlled nonlinear systems
Chemical Engineering Science, 2004, 59(20), 4325-4338
Abstract:
This paper presents an approach to analyzing robustness properties of nonlinear
systems under feedback control. The core idea is to apply numerical bifurcation
analysis to the closed-loop process, using the controller/observer tuning parameters,
the set points, and parameters describing model uncertainty (parametric as well
as unmodeled dynamics) as bifurcation parameters. By analyzing the Hopf bifurcation
and saddle-node bifurcation loci with respect to these parameters, bounds on the
controller tuning are identified which can serve as a measure for the robustness
of the controlled system. These bounds depend upon the type as well as the degree
of mismatch that exists between the plant and the model used for controller design.
The method is illustrated by analyzing three control systems which are applied to a continuously operated stirred tank reactor: a state feedback linearizing controller and two output feedback linearizing controllers. While model uncertainty has only a minor effect on the tuning of the state feedback linearizing controller, this does not represent a very realistic scenario. However, when an observer is implemented in addition to the controller and an output feedback linearizing scheme is investigated, it is found that the plant-model mismatch has a much more profound impact on the tuning of the observer than it has on the controller tuning. In addition, two observer designs with different level of complexity are investigated and it is found that a scheme which makes use of additional knowledge about the system will not necessarily result in better stability properties as the level of uncertainty in the model increases. These investigations are carried out using the robustness analysis scheme introduced in this paper.
Keywords:
Nonlinear system; Bifurcation analysis; Feedback linearizing control; Robust controller tuning