Input-Output Polynomial Models

Input-output polynomial models, including ARX, ARMAX, output-error, and Box-Jenkins model structures

A polynomial model uses a generalized notion of transfer functions to express the relationship between the input, u(t), the output y(t), and the noise e(t) using an equation of the form:

$A (q) y (t) = \frac{B (q)}{F (q)} u (t - n k) + \frac{C (q)}{D (q)} e (t) .$

A(q), B(q), F(q), C(q) and D(q) are polynomial matrices in terms of the time-shift operator q^-1. u(t) is the input, and nk is the input delay. y(t) is the output and e(t) is the disturbance signal.

Each polynomial has an independent order, or number of estimable coefficients. For example, if A(q) has an order of 2, then the A polynomial has the form A(q) = 1 + a₁q^-1 + a₂q^-2.

In practice, not all the polynomials are simultaneously active. Simpler polynomial forms, such as ARX, ARMAX, Output-Error, and Box-Jenkins provide model structures suitable for specific objectives such as handling nonstationary disturbances or providing completely independent parameterization for dynamics and noise. For more information about these model types, see What Are Polynomial Models?

Apps

System Identification

Identify models of dynamic systems from measured data

Functions

expand all

Create Polynomial Model

`idpoly`	Polynomial model with identifiable parameters
`arx`	Estimate parameters of ARX, ARIX, AR, or ARI model
`armax`	Estimate parameters of ARMAX, ARIMAX, ARMA, or ARIMA model using time-domain data
`bj`	Estimate Box-Jenkins polynomial model using time-domain data
`iv4`	ARX model estimation using four-stage instrumental variable method
`ivx`	ARX model estimation using instrumental variable method with arbitrary instruments
`oe`	Estimate output-error polynomial model using time-domain or frequency-domain data
`polyest`	Estimate polynomial model using time- or frequency-domain data
`pem`	Prediction error minimization for refining linear and nonlinear models

Find Best ARX Model-Order Combination

`arxstruc`	Compute loss functions for single-output ARX models
`ivstruc`	Compute loss functions for sets of ARX model structures using instrumental variable method
`selstruc`	Select model order for single-output ARX models
`struc`	Generate model-order combinations for single-output ARX model estimation

Model Initialization and Structure Parameters

`arxRegul`	Determine regularization constants for ARX model estimation
`delayest`	Estimate time delay (dead time) from data
`init`	Set or randomize initial parameter values

Extract or Set Model Parameters

`polydata`	Access polynomial coefficients and uncertainties of identified model
`getpvec`	Obtain model parameters and associated uncertainty data
`setpvec`	Modify values of model parameters
`getpar`	Obtain attributes such as values and bounds of linear model parameters
`setpar`	Set attributes such as values and bounds of linear model parameters
`setPolyFormat`	Specify format for B and F polynomials of multi-input polynomial model

Estimation Options

`armaxOptions`	Option set for `armax`
`arxOptions`	Option set for `arx`
`arxRegulOptions`	Option set for `arxRegul`
`bjOptions`	Option set for `bj`
`iv4Options`	Option set for `iv4`
`oeOptions`	Option set for `oe`
`polyestOptions`	Option set for `polyest`

Topics

Polynomial Model Basics

What Are Polynomial Models?
Polynomial model structures including ARX, ARMAX, output-error, and Box-Jenkins.
Data Supported by Polynomial Models
Use time-domain and frequency-domain data to estimate discrete-time and continuous-time models.

Estimate Polynomial Models

Preliminary Step – Estimating Model Orders and Input Delays
To estimate polynomial models, you must provide input delays and model orders.
Estimate Polynomial Models in the App
Import data into the app, specify model orders, delays and estimation options.
Estimate Polynomial Models at the Command Line
Specify model orders, delays, and estimation options.
Polynomial Sizes and Orders of Multi-Output Polynomial Models
Size of A, B, C, D, and F polynomials for multi-output models.
Estimate Models Using armax
This example shows how to estimate a linear, polynomial model with an ARMAX structure for a three-input and single-output (MISO) system.

Set Polynomial Model Options

Specifying Initial States for Iterative Estimation Algorithms
Specify initial conditions of polynomial models for iterative estimation algorithms.
Polynomial Model Estimation Algorithms
Choose between the ARX and IV algorithms for ARX and AR model estimation.

Featured Examples

Spectrum Estimation Using Complex Data - Marple's Test Case

Perform spectral estimation on time series data. We use Marple's test case (The complex data in L. Marple: S.L. Marple, Jr, Digital Spectral Analysis with Applications, Prentice-Hall, Englewood Cliffs, NJ 1987.)

Open Live Script

Picking Instrumental Variables for System Identification

Illustration of the meaning and choice of instrumental variables (IVs) for system identification.

Open Live Script