Skip to Main Content
This paper presents a new method of stabilizing closed-loop systems, whereby instability warnings and overriding controls are performed. The system stability is maintained by having a separated agent, here called stabilizing agent, to produce corrective actions to the manipulated variables. This agent computes the stabilizing thresholds on-the-fly based on dissipative projections and real-time stability conditions. It overrides the controller outputs with stabilizing values once the threshold is violated. The plant models are not employed in this approach. Instead of that, online information of manifest variables is utilized to evaluate necessary bounds on manipulated variables for the stability purpose. Theoretical results are illustrated in simulation of a model predictive controller for an unit operation of the paper mill.