Specifications

Control Based On Multiple Plant Models

4-15

Control Based On Multiple Plant Models

The “Nonlinear CSTR Application” on page 4-9 shows how updates to the prediction

model can improve MPC performance. In that case the model is nonlinear and you can

obtain frequent updates by linearization.

A more common situation is that you have several linear plant models, each of which

applies at a particular operating condition and you can design a controller based on

each linear model. If the models cover the entire operating region and you can define a

criterion by which you switch from one to another as operating conditions change, the

controller set should be able to provide better performance than any individual controller.

The Model Predictive Control Toolbox includes a Simulink block that performs this

function. It is the Multiple MPC Controllers block. The rest of this section is an

illustrative example organized as follows:

• “A Two-Model Plant” on page 4-15

• “Designing the Two Controllers” on page 4-17

• “Simulating Controller Performance” on page 4-18

A Two-Model Plant

Note Switching MPC Controllers with Multiple MPC Controllers Block provides an

animated version of the plant described below.

Animation of the Multi-Model Example is a stop-action snapshot of the subject plant. It

consists of two masses, M

and M

. A spring connects M

to a rigid wall and pulls it to the

right. An applied force, shown as a red arrow in Animation of the Multi-Model Example,

opposes this spring, pulling M

to the left.

When the two masses are detached, as in Animation of the Multi-Model Example, mass

is uncontrollable and responds only to the spring pulling it to the left.

If the two masses collide, however, they stick together (the collision is completely

inelastic) until a change in the applied force separates them.

The control objective is to move M

in response to a command signal. The blue triangle

in Animation of the Multi-Model Example represents the desired location. At the instant

shown, the desired location is –5.