Case Study Problems in Multivariable Control

Some real multivariable process control problems are presented here. Follow the links to the case study problem of your interest. Each of these problems deals with the control of (typically) a 2x2 MIMO system. Some of the case studies do have more than two outputs, in which case you are required to control two outputs and monitor the performance of the others.

Each project will, in general, involve the following phases : literature review, model development and process identification, control structure selection and controller tuning for SISO systems, multiple SISO loop tuning. Details will be provided by your project 'advisor' at the appropriate time. Some general guidelines are provided below, and deescriptions of the individual projects may be obtained by following the links.

A Simulink file appropriate to your case study will be provided by your project advisor, and will play the role of the 'plant' you want to control. While developing your model for the process, bear in mind that the output of the Simulink file contains the 'true' output of the plant corrupted by (measurement) noise.

Anaerobic Sludge Digesters : Anaerobic sludge digesters are used to degrade compounds (in waste waters high in suspended solids) to carbon dioxide and water.

Drug Infusion System : This deals with the administration of drugs to a patient undergoing surgery or some sort of trauma (e.g., a heart attack)

Fluidized Catalytic Cracking Unit : A fluidized catalytic cracking unit (FCCU) is used to produce the majority of gasoline in an oil refinery.

Reactive Ion Etcher : Plasma etching (or reactive ion etching) is an important process in the semiconductor device manufacturing.

Rotary Lime Kiln Temperature Control : Lime kilns are used in the paper (and other) industries to convert lime mud to lime.

GENERAL GUIDELINES

The following are important issues to consider:

What are the important process disturbances?

Do the orders of magnitudes of the gains, time constants and time-delays make physical sense?

If using independent SISO control loops (MVSISO), what pairings are suggested by a Relative Gain Array (RGA) analysis?

If an SISO controller is tuned based on the other loops being open, does the tuning need to change when all of the other loops are closed?

Should a decoupling (steady-state or dynamic) type of control strategy be used, for real (MIMO) multivariable control? What about multivariable IMC?

Can you think of any other measured or manipulated variables that could be used?

When tuning the independent SISO loops (assuming the other loops are open) it is probably best to use a procedure that you are familiar with such as IMC-based PI, etc. You will generally use a closed-loop time constant (l, the IMC filter factor) of roughly one-half of the dominant open-loop time constant (if this satisfies the minimum requirements for l based on the deadtime).

When testing controller performance for these multivariable processes, it is probably best to initially make a step change in one setpoint while keeping all other setpoints constant (that is, change r1 without changing r2, etc.). Do this for each loop. Compare how a setpoint change in one output variable affects the other output variables. Also consider the manipulated variable responses.