System Identification Toolbox 7.3.1

Parametric Model Estimation

This narrated video demonstration provides an overview of the System Identification Toolbox and shows how to estimate a linear model of a dynamic system using the product's graphical user interface.

The demonstration segments run for a total of 3:30 minutes. Please use the controls at the bottom of each demonstration window to control playback.

Contents

Overview

The System Identification Toolbox provides a graphical user interface that lets you:

Figure1: This is the System Identification Toolbox GRaphical User Interface

View a demonstration of general System Identification Toolbox features (30 seconds)

Importing and Moving Data Sets

The toolbox lets you import data from a MATLAB data file (.mat) or from an IDDATA object file, a format unique to the toolbox. In this demonstration, we will import two data files, one for estimating the model and a second for validating the model.

View a demonstration on importing and manipulating data sets (50 seconds)

Preprocessing Data

Data sometimes contains disturbances and offsets. The toolbox provides features to view the data and remove these artifacts to help ensure proper model identification.

View a demonstration on preprocessing data (75 seconds)

Estimating and Validating Models

The toolbox offers both parametric and nonparametric methods for estimating a model. In this demonstration, we will use parametric models. We will then evaluate multiple estimated models against the validation data set.

View a demonstration on estimating and validating models (50 seconds)