Cannot reproduce MIMO StateSpace result. about sippy HOT 2 CLOSED

Dear Gopal,

You can choose the correct order with 'SS_fixed_order=3' .

The identification proposed in the paper is closed-loop, therefore it is different from that simulated by you.

You have small typo errors in the code as follows:

"fset.white_noise_var: this function generates a white noise matrix (rows with zero mean). The input arguments are:

1. L: size (number of columns of the returned matrix);
2. Var: variance list;
   e.g. the following command line:
   noise=white_noise_var(100,[1,2])
   returns a matrix that has 100 columns and two-row vectors with variances 1 and 2"

Var= [0.15, 0.15] # you have 2 outputs
noise = fset.white_noise_var(npts, Var)

Consider that lower variances are used in the paper.

"x,y_hat=SS_lsim_innovation_form(A,B,C,D,K,y,u,x0)"

xid, yid = fsetSIM.SS_lsim_innovation_form(sys_id_p.A, sys_id_p.B,
sys_id_p.C, sys_id_p.D,
sys_id_p.K, y_tot, U, sys_id_p.x0)

The order of U and Y is reversed.

Hoping to have answered your questions.
Best Wishes, the Sippy teams.

from sippy.

Hi SIPPY teams,

Thank you for your response. After making the changes you mentioned, the script works as expected. Appreciate the help. I close this issue now.

Regards,
Gopal

from sippy.