| System Identification Toolbox | |
Transformations to Other Model Representations
Within the structure in which the model was created, you can extract parametric information by get or by subscripting. For example, for a state-space model, Mod.A is the A-matrix, while Mod.dA contains its standard deviations. For a polynomial model, Mod.a and Mod.da are the A-polynomial and its standard deviation, etc.
In addition, regardless of the particular model structure, there are a number of commands that compute various model representations. These all have the basic syntax
where G contains model characteristics and dG their standard deviation or covariance. The transformation commands are
[A,B,C,D,K,X0,dA,dB,dC,dD,dK,dX0] = ssdata(Model) [a,b,c,d,f,da,db,dc,dd,df] = polydata(Model) [A,B,dA,dB] =arxdata(Model) [Num,Den,dNum,dDen] = tfdata(Model) [Z,P,K,CovZ,CovP,covK] = zpkdata(Model) G = idfrd(Model) [H,w,CovH] = freqresp(Model)
The two last commands were described previously. The three first commands clearly transform to the state-space, the polynomial, and the multivariable ARX representations. See Defining Model Structures. tfdata and zpkdata compute the transfer functions and zeros, poles and transfer function gains. See the reference pages in Chapter 4 for details.
| | Graphs of Model Properties | Discrete and Continuous Time Models | |