Process Control Challenge

Process Control Challenge Problems

updated 2024-02-04

We use simulators to test many aspects of control and automation algorithms.  Desirably, the challenge problems should be simple to understand and implement, as well, they should credibly represent issues characteristic of the applications.  Control in the chemical process industries (CPI) has a confluence of difficulties that are different from the mechatronic test cases or the deterministic ODEs that characterize many investigations.  This article Process Control Challenge Problems 2018-11-15 discusses such issues and how to assess controller performance for the CPI.

The first four of these process control challenges are the case study challenges in my book “Nonlinear Model-Based Control using First-Principles Models in Process Control” to be published early 2024 by ISA (the International Society of Automation).  Visit https://www.isa.org/standards-and-publications/isa-publications/isa-books.

  1. This link Hot & Cold Mixing Process Simulator Description 2018-09-20 describes an in-line hot and cold water mixing process.  It has two controlled variables (total flow and mixed fluid temperature) and two manipulated variables (signals to the hot and cold valves).  It is nonlinear and interactive, subject to constraints, and the best pairing changes with temperature setpoint.  It currently has two control options SISO PIDs and a simple model based controller.  It is an excellent test case to explore SISO pairing, ratio, decouplers, gain scheduling, and constraint handling.  And, this link Hot and Cold Mixing 2018-09-17 provides access to my water mixing simulator with user options to use either PID or a model-based controller.  Of course, you can include your own controller.
  2. This link pH Modeling and Simulator Description 2021-10-02 describes a pH neutralization process.  It also provides user directions for the simulator (accessed by the link below).  It has two controlled variables, in-tank pH and tank level, and two manipulated variables, caustic inflow and discharge rate.  If the pH is out of discharge limits, then the tank discharge is stopped, and inflow accumulates until pH is back under control.  Then discharge control seeks to return tank contents to the set point level.  It is a nonlinear and non-stationary process with an override, and may be very applicable algorithms seeking to manage these issues through gain scheduling, nonlinear control, or adaptive control.   This link pH Process Simulator w Heuristic Control 2021-10-02 provides access to the pH neutralization simulator.  The current controller is a very simple set of heuristic rules as a placeholder for you to add a better control algorithm.  The process is validated by several publications, including: Choi, J. Y., H. G. Pandit, R. R. Rhinehart, and R. J. Farrell, “A Process Simulator for pH Control Studies,” Computers & Chemical Engineering, Vol. 19, No 5, pp. 527-540, (1995); and Choi, J. Y., H. G. Pandit, and R. R. Rhinehart, “A Process Simulator for pH Control Studies,” 1993 ACC Conference Proceedings, San Francisco, CA, June, 1993 paper FM9-11:40, pp. 2594-5.
  3. This simulator r3eda PMBC Car Speed Control LF to Solve for u implicit 2017-04-23 provides a SISO case based on an automobile speed control.  The single controlled variable is vehicle speed, and the single manipulated variable is accelerator pedal position.  Although SISO and with a relatively simple model, the process is nonlinear and subject to environmental disturbances and constraints.  User instructions are included with the VBA code.
  4. This heat exchanger simulator uses hot liquid to heat cold liquid in a counter-current heat exchanger.  It is nonlinear, and has a moderately high order.  The deadtime, which is primarily due to transport delay, depends cold fluid flow rate.  It would be useful in demonstrating issues in process control, nonlinear and adaptive algorithms, and cascade.  Here HX Dynamic Simulator w PID 2021-09-09 provides access to the simulator.  And here ACC 2022 HX Simulator 2021-09-14 is a link to download the process description.  This is based on a publication, Rhinehart, R. R., “A Heat Exchanger Simulator for Testing and Demonstrating Automation Algorithms”, Proceedings of the 2022 American Control Conference, Atlanta, GA, June 7-10, Paper FrA12.6.
  5. This challenge problem is mixing in a CST to maintain both outlet composition and tank level.  A concentrated titrant is added to a wild flow with time-varying composition (unmeasured) and flow rate.  The outlet flow rate controls level.  The basic equations for the dynamic process are simple to understand and implement, but dead zones in the mixer and calibration errors on measurements are some confounding effects.  The time-constant for composition (first-order-ish) but depends on flow rate and tank level (which is integrating).  The process is nonlinear, nonstationary, interactive, noisy, and subject to constraints on flow rates and level.  The simulator is here PMBC-CST-Level-and-Composition-2024-02-02.xlsm, and the description and user guide is here Equation-Detail-for-Simple-MBC-Tank-Level-and-Composition-2024-02-02-1.docx.

I anticipate adding a flash tank and a fed-batch process simulator in the near future.