Assessment of dynamic matrix control controller parameters via estimating Markov parameters of disturbance

Abstract

For an industrial control system, controller parameters are important factors that affect the control system performance. This article introduces a controller parameter performance assessment method based on disturbance characteristic variation for a dynamic matrix control system. Assuming that the process model is accurate and the setpoint is constant, disturbance characteristic variation without adjusting controller parameters leads to the degradation of system performance. Hence, the Markov parameters of a disturbance model, which inflect the variation of disturbance characteristics, are used to assess the controller parameter performance. These Markov parameters can be obtained from closed-loop data using the subspace identification method. The differences between the Markov parameters in an actual running state and those in a well-regulated state are used to design the assessing index of controller parameters. The simulation results in the Wood-Berry model and TE process show the validity of the proposed index.