Disturbance Rejection Tuning of a State-Space Predictive Controller for a Gas Conditioning Unit

This paper presents an experimental application of a state-space predictive controller. The controller is used for control of a gas conditioning unit, an integrating process with two inputs and two outputs, in an industrial environment. The predictive controller is used as a general-purpose low-level controller, which is not usual for the established practice of industrial model predictive control (MPC). In this role, efficient disturbance rejection performance is required. For the considered process the conventional output step disturbance model results in a slow feedback response, and is also structurally inappropriate. An original approach to the tuning of the feedback action based on the Kalman filter is introduced. The disturbance model is suitable for integrating and unstable processes. Robustness analysis using nominal sensitivity functions and a set of possible true models is provided. The controller was applied successfully, despite some real-life problems during commissioning. Feedback performance and also the robustness indicators were improved in comparison to the standard control. Key-Words: Disturbance rejection; Predictive control; Kalman filter; Feedback systems; Robustness analysis