System Identification Toolbox

The Toolbox Commands

It may be useful to recognize several layers of the System Identification Toolbox. Initially concentrate on the first layer of basic tools, which contains the commands from the System Identification Toolbox that any user must master. You can proceed to the next levels whenever an interest or the need from the applications warrants it. The layers are described in the following paragraphs:

Layer 0: Help Functions.   Help ident gives an overview of available commands. idhelp gives access to a "micromanual" of command-line help, with several subhelps like idhelp, evaluate, etc.

Layer 1: Basic Tools for Estimating Black-Box Models.   The first layer contains the basic tools for estimating models from measured data. It is necessary to know the basics of the data representation and the simple commands to build and evaluate black-box models. The commands are:

• For data representation: iddata, plot
• For nonparametric estimation of impulse and frequency response: impulse, step, spa
• For estimating black-box models of state-space and input-output type: pem, arx
• For evaluating models: compare, resid
• For displaying model characteristics: bode, nyquist, pzmap, step, view
• Looking at parametric model characteristics: By field referencing, like Mod.A, Mod.dA, etc.

The corresponding background is given in the next few sections of this "Tutorial."

Layer 2: Creating Models for Simulation and Transforming Models.   To define models, to generate inputs and simulate models

• idarx, idpoly, idss, idinput, sim
  

To transform models to other representations

• arxdata, polydata, ssdata, tfdata, zpkdata
  

Layer 3: Model Structure Selection.   The third layer of the toolbox contains some useful techniques to select orders and delays.

• arxstruc, selstruc
  

Layer 4: Structured Models and Further Model Conversions.   The fourth layer contains transformations between continuous and discrete time, and functions for estimating completely general model structures for linear systems. The commands are

• c2d, d2c, idss, idgrey, pe, predict
  ss, tf, zp, frd (to be used with the Control System Toolbox)
  

The corresponding material is covered in Defining Model Structures and in Examining Models.

Layer 5: Recursive Identification.   Recursive (adaptive, online) methods of parameter estimation are covered by the commands

• rarmax, rarx, rbj, roe, rpem, rplr
  

They are covered in Recursive Parameter Estimation.

(See the beginning of the Function Reference section for a complete list of available functions.)

Tutorial

An Introductory Example to Command Mode