|
Hello everybody,
I try to use the robust control toolbox to create a H_inf controller. For the beginning I want to do it without the weights, which you need for the mixed problem.
At first, I create a simple plant G and then a generalized plant P with sysic. The goal is, to minimize H_inf norm for the influence of the input w on z (that's is the error signal -> input for controller). But as result, I get a controller with nearly zero magnitude. That means, the error doesn't decrease over time. I consider a simple standard close loop without any disturbances.
Where is my fault?
Below my code:
%% generalized plant
G = tf([10],[1 1]);
systemnames = 'G'
inputvar = '[r ; u]'
input_to_G = '[u]'
outputvar = '[r-G ; r-G]'
sysoutname = 'P'
sysic;
[K,CL, gopt] = hinfsyn(P,1,1);
Thx!
|