User`s guide

4 Nonlinear Black-Box Model Identification

Nonlinearity Object Name Supported Model Type Supports Multiple

Inputs?

Tree Partition

treepartition

Nonlinear ARX

Yes

Wavelet Network

wavenet

Hammerstein-Wiener and

Nonlinear ARX

Yes

The Neural Network nonlinearity lets you import a n etwork object you

created using the Neural Network Toolbox commands.

The nonlinearity estimators

deadzone, poly1d, pwli near ,andsat ura tion

are optimized for estimatin g Hammerstein-Wiener models. However, you

can also these estimators with nonlinear A RX models that have only one

regressor. For more information about n onlinear A RX model structure, see

“Deﬁnition of the N o nlinear ARX Model” on page 4-4.

Creating Custom Nonlinearities

You can create custom nonlinearities for nonlinear ARX and

Hammerstein-Wiener models.

A custom nonlinearity uses a unit function that you d eﬁne. This custom unit

function uses a weighted sum of inputs to compute a scalar output.

You can use a combination of these unit functions to approximate the

nonlinearity.

Note Hammerstein-Wiener models require that your custom nonlinearity

have one input and one output.

function [f, g, a] = gaus suni t(x)

%GAUSSUNIT exam ple of customnet unit function

%[f, g, a] = GAUSSUNIT(x)

% x: unit function variable

% f: unit function value

4-26