This is machine translation

Translated by Microsoft
Mouseover text to see original. Click the button below to return to the English verison of the page.

Note: This page has been translated by MathWorks. Please click here
To view all translated materals including this page, select Japan from the country navigator on the bottom of this page.

Simulating Model Predictive Controller with Plant Model Mismatch

This example shows how to simulate a model predictive controller under a mismatch between the predictive plant model and the actual plant.

The predictive plant model has 2 manipulated variables, 2 unmeasured input disturbances, and 2 measured outputs. The actual plant has different dynamics.

Define Plant Model

Define the parameters of the nominal plant which the MPC controller is based on. Systems from MV to MO and UD to MO are identical.

p1 = tf(1,[1 2 1])*[1 1; 0 1];
plant = ss([p1 p1],'min');
plant.InputName = {'mv1','mv2','ud3','ud4'};

Design MPC Controller

Define inputs 1 and 2 as manipulated variables, 3 and 4 as unmeasured disturbances.

plant = setmpcsignals(plant,'MV',[1 2],'UD',[3 4]);
% Create the controller object with sampling period, prediction and control
% horizons:
mpcobj = mpc(plant,1,40,2);
-->The "Weights.ManipulatedVariables" property of "mpc" object is empty. Assuming default 0.00000.
-->The "Weights.ManipulatedVariablesRate" property of "mpc" object is empty. Assuming default 0.10000.
-->The "Weights.OutputVariables" property of "mpc" object is empty. Assuming default 1.00000.

For unmeasured input disturbances, the MPC controller will use the following unmeasured disturbance model.

distModel = eye(2,2)*ss(-.5,1,1,0);
mpcobj.Model.Disturbance = distModel;

Define the Real Plant Model Used in Simulation

Define the parameters of the actual plant in closed loop with the MPC controller.

p2 = tf(1.5,[0.1 1 2 1])*[1 1; 0 1];
psim = ss([p2 p2],'min');
psim = setmpcsignals(psim,'MV',[1 2],'UD',[3 4]);

Simulate Closed-Loop Response Using the SIM Command

Define reference trajectories and unmeasured disturbances entering the actual plant.

dist = ones(1,2);   % unmeasured disturbance signal
refs = [1 2];       % output reference signal
Tf = 20;            % total number of simulation steps

Create an MPC simulation object.

options = mpcsimopt(mpcobj);
options.unmeas = dist;  % unmeasured disturbance signal
options.model = psim;   % real plant model

Run the closed-loop MPC simulation with model mismatch and unforeseen unmeasured disturbance inputs.

-->Converting model to discrete time.
-->Assuming output disturbance added to measured output channel #1 is integrated white noise.
-->Assuming output disturbance added to measured output channel #2 is integrated white noise.
-->The "Model.Noise" property of the "mpc" object is empty. Assuming white noise on each measured output channel.
-->Converting model to discrete time.

The closed loop tracking performance is acceptable with the presence of unmeasured disturbances.

Was this topic helpful?