Scilab Home page | Wiki | Bug tracker | Forge | Mailing list archives | ATOMS | File exchange
Please login or create an account
Change language to: English - Français - Português - 日本語

Please note that the recommended version of Scilab is 6.0.0. This page might be outdated.
See the recommended documentation of this function

Справка Scilab >> CACSD > arma

arma

Scilab arma library

Description

Armax processes can be coded with Scilab tlist of type 'ar'. armac is used to build Armax scilab object. An 'ar' tlist contains the fields ['a','b','d','ny','nu','sig'].

armac

this function creates a Scilab tlist which code an Armax process A(z^-1)y= B(z^-1)u + D(z^-1)sig*e(t)

-->ar=armac([1,2],[3,4],1,1,1,sig);
 
-->ar('a')   
 ans  =
 
!   1.    2. !
-->ar('sig')
 ans  =
 
    1.
armap(ar [,out])

Display the armax equation associated with ar

armap_p(ar [,out])

Display the armax equation associated with ar using polynomial matrix display.

[A,B,D]=armap2p(ar)

extract polynomial matrices from ar representation

armax

is used to identify the coefficients of a n-dimensional ARX process A(z^-1)y= B(z^-1)u + sig*e(t)

armax1

armax1 is used to identify the coefficients of a 1-dimensional ARX process A(z^-1)y= B(z^-1)u + D(z^-1)sig*e(t)

arsimul

armax trajectory simulation.

narsimul

armax simulation ( using rtitr)

odedi

Simple tests of ode and arsimul. Tests the option 'discret' of ode

prbs_a

pseudo random binary sequences generation

reglin

Linear regression

Example

// Example extracted from the demo arma3.dem.sce in the cacsd module
// Spectral power estimation 
// ( form Sawaragi et all) 
m = 18;
a = [1,-1.3136,1.4401,-1.0919,+0.83527];
b = [0.0,0.13137,0.023543,0.10775,0.03516];
u = rand(1,1000,'n');
z = arsimul(a,b,[0],0,u);

//----Using macro mese 
[sm,fr]=mese(z,m);

//----The theorical result 

function gx=gxx(z, a, b)
  w  = exp(-%i*2*%pi*z*(0:4))'
  gx = abs(b*w)^2/(abs(a*w)^2);
endfunction

res=[];
for x=fr
  res=[ res, gxx(x,a,b)];
end

//----using armax estimation of order  (4,4)
// it's a bit tricky because we are not supposed to know the order

[arc,la,lb,sig,resid]=armax(4,4,z,u);
res1=[];
for x=fr
  res1=[ res1, gxx(x,la(1),lb(1))];
end

//-- visualization of the results
plot2d([fr;fr;fr]',[20*log10(sm/sm(1));20*log10(res/res(1));20*log10(res1/res1(1))]',[2,1,-1])
legend(["Using macro mese";"Theoretical value";"Arma identification"])
xtitle("Spectral power","frequency","spectral estimate")
Scilab Enterprises
Copyright (c) 2011-2017 (Scilab Enterprises)
Copyright (c) 1989-2012 (INRIA)
Copyright (c) 1989-2007 (ENPC)
with contributors
Last updated:
Thu Oct 02 14:01:06 CEST 2014