Track: Modeling and Simulation

Abstract

Proportional-integral-derivative (PID) controllers are widely used in industrial processes, and finding the most appropriate PID parameters plays an important role in optimizing plant operation. However, tuning a large number of PIDs is a labor-intensive task that often leaves many control engineers uncertain about how to proceed. To tackle this problem, this article provides an automated approach to select and recommend the appropriate set of numerical values for P, I, and D. Our approach is an open-loop PID autotuner which is based on a data-driven model that represents the physical process and an ensemble of optimization algorithms. To evaluate the effectiveness of our approach, we tuned four PIDs that are responsible for controlling multiple-effect evaporators, a critical piece of equipment in kraft pulp manufacturing used to evaporate and concentrate black liquor. The results show clear success in tuning several PIDs automatically with basic process knowledge and minimum effort.