- To analyse a system with multiple control loops using the RGA method and hence decide on the optimal method of matching up the measured variables with manipulated variables
- To see the effect of differing experimental conditions in the above analysis
- To run the system with two control loops in place to see the interaction of the two control loops

Module 6 - Multivariable Systems.

- A tank with two inputs and one output:
- Input 1 is a concentrate stream
- Input 2 is a diluent stream
- Output depends on level in tank

- Three controllers:
- Controller 1 (with lag) controlling the level
- Controller 2 (with delay) controlling the concentration
- Controller 3 is a flow controller on one stream ie one flowrate is constant

There are two different ways of running the above system.

- With a constant outflow
- With a constant diluent flow

The experimental setup and procedure for each is discussed in turn below.

- Controller 3 is placed on the outflow from the tank so that there is a constant valve constant and the outflow depends only on the level in the tank.
- Manipulated variable 1 is the concentrate flow
- Manipulated variable 2 is the diluent flow
- Measurement 1 is the concentrate concentration
- Measurement 2 is the level

The experimental apparatus therefore is shown below:

The aim of the experiment is to try and match up the measurements and manipulations using the RGA method discussed in the theory section. This will depend in part on the initial conditions in the tank and the operating conditions during the experiment. Therefore there are two different experiments.

Following the links below will take you to a further discussion about the experimental conditions along with pre-experimental questions to guide you through each simulation.

- Controller 3 is placed on the diluent flow which stays constant throughout.
- Manipulated variable 1 is the concentrate flow
- Manipulated variable 2 is the flow out valve constant
- Measurement 1 is the concentrate concentration
- Measurement 2 is the level

The experimental apparatus again is shown. Note that this time there is only one way to run this experiment.

The first aim of this experiment is to evaluate the gain matrix and RGA and show that there is only one way to match up these variables (the flow out from the tank does not have any effect on the concentration in the tank). The second aim of the experiment is to run the two loops separately and then run them together to see how they interact and how the tuning is affected.

Following the link below will take you to a further discussion about the experimental conditions along with pre-experimental questions to guide you through the simulation.

Return to the Virtual Control Laboratory