convert ARMAX to discrete or continiuous transfer function

hi here is what i have found as armax estimation

>> amx2221

amx2221 =
Discrete-time ARMAX model:  A(z)y(t) = B(z)u(t) + C(z)e(t)

    A(z) = 1 - 1.12 z^-1 + 0.1199 z^-2                      

    B(z) = 0.0004663 z^-1 - 0.0004681 z^-2                  

    C(z) = 1 - 1.154 z^-1 + 0.1929 z^-2                     

Name: amx2221
Sample time: 1.5e-05 seconds

Parameterization:
   Polynomial orders:   na=2   nb=2   nc=2   nk=1
   Number of free coefficients: 6
   Use "polydata", "getpvec", "getcov" for parameters and their uncertainties.

Status:                                           
Estimated using POLYEST on time domain data "p50".
Fit to estimation data: 98.35% (prediction focus) 
FPE: 8.796e-12, MSE: 8.791e-12

while i use tf(amx2221) it answers me:

>> tf(amx2221)

ans =

From input "u1" to output "y1":
0.0004663 z^-1 - 0.0004681 z^-2
-------------------------------
  1 - 1.12 z^-1 + 0.1199 z^-2

Name: amx2221
Sample time: 1.5e-05 seconds
Discrete-time transfer function.

or in the other shape:

>> d2c(tf(amx2221))

ans =

    From input "u1" to output "y1":
      75.11 s - 1.904e04
    -----------------------
    s^2 + 1.414e05 s - 9996

Continuous-time transfer function.

and this is exactly B(z)/A(z) and there is no effect of C(z);

How can i get a transfer function that have the effect of C(z)?

Answer 1

[vidya kurada]

tf(amx2221,'noise') should give the noise model