Account for Uncertainty with Robust Control Design: Part 1
Robust control design has been increasingly used in industrial settings by leading automation companies. The design procedure has evolved in the last decades and fairly automated procedures exist now for use by practicing engineers or even operators. One does not need to be familiar with the details of the underlying theory to use it. Robust control is different than conventional control in that it accounts for uncertainty bounds and designs a controller with known/desired performance and stability characteristics. Robust control can be applied to multivariable or Single Input Single Output (SISO) processes. This paper is aimed at providing a tutorial on the Robust PID control design approach to practicing chemical engineers. We use the classical pH control problem as an example, which is a challenging problem due to its non-linearity. First, we analyze the pH process by using the benchmark model of Henson and Seborg. We identify the fundamental limitations of the linear control design in terms of model uncertainty and sensor sampling constraints. Subsequently, we design a controller following the guidelines from robust control theory. Finally, we demonstrate the results though implementation in a lab-scale wastewater system. The experimental results show the validity of the process model and the control design approach. It also points out the limitations of the linear controller performance, leading to an interesting follow-up work regarding gain scheduling and adaptation.
- Author (aut): Josh, Rakesh
- Author (aut): Tsakalis, Konstantinos
- Author (aut): MacArthur, J. Ward
- Author (aut): Dash, Sachi
- Contributor (ctb): Ira A. Fulton Schools of Engineering