MATLAB: Use N4sid for system identification (for new data estimation)

MATLABmatlab system identificationSystem Identification Toolbox

Hi,

I am using N4sid toolbox to do a system identification problem. I used the command: "m = n4sid(z, 5,'Form','free', opt);" to create the system based on measured "input" and "output" (contains in z=z = iddata(y', u', 0.2, 'Name', 'heating/cooling load', 'TimeUnit', 's');".

However,I still want to use this model to predict some unmeasured cases. To be more specific, I have a new input measurement U2, I want to used to model to predict the output from U2. How can I realize that? What is command for this estimation using the model from N4sid?

Thanks very much.

Best Answer

Computing response of a model to a given input is called simulation. Use the idParametric/sim command for that as in:

 Ud = iddata([], U2, .2);
 Yd = sim(m, Ud);
 Y2 = Yd.y; % the raw response data

See:

http://www.mathworks.com/help/ident/ref/sim.html

http://www.mathworks.com/help/ident/ref/findstatesidparametric.html

Related Solutions

MATLAB: How to export data/transfer functions from system identification to Matlab workspace using script

Suppose you have testit.sid then

S = load('testit.sid', '-mat');

Now S will be a field with several fields.

The field named Data will (probably) be a cell array of iddata, one entry for each square that is filled in in the "Import data" section.

The field named Models will be a cell array of various data types, one entry for each square that is filled in in the "Import models" section.

For example in my test, S.Models{2} is a 1 x 2 idtf . Its parts can be examined with, for example, S.Models{2}.Numerator and S.Models{2}.Denominator and S.Models{2}.Report is a struct with information about the transfer function estimation, and for example, S.Model{2}.Report.Parameters.Labels and S.Model{2}.Report.Parameters.ParVector

MATLAB: System identification procedure

Have Simulink data available in MATLAB workspace as double vectors. In the GUI (ident) import time signals to load the data. Then you work from there. If you are estimating models in the command line, package the data into an iddata object:

 data = iddata(y,u,Ts); % y = output; u = input; Ts = sample time

Then try various estimation commands, e.g:

 model = tfest(data, 2); % second order transfer function

The logged data (y,u) must be equi-sampled. If not, interpolate them to a uniform time grid.

Best Answer

Related Solutions

MATLAB: How to export data/transfer functions from system identification to Matlab workspace using script

MATLAB: System identification procedure

Related Question