Datatype error nlgreyest parameters

I would appreciate any help on resolving the error
Error using iddata/nlgreyest
In 'Ident:estimation:infeasibleSimulation', data type supplied is incorrect for parameter {0}.
Error in identifyModel (line 64)
model = nlgreyest(dataObject, model, opt);
I assume parameter {0} means it concerns the first parameter, although matlab typically doesn't use zero indexing. I use the following code to create my parameter struct for idnlgrey, and all values are standard doubles.
function identificationParam = defineParamModel()
a11 = -1;
a22 = -1;
a33 = -1e-1;
a23 = 0;
a32 = 0;
identificationParam(1).Name = 'dynamic matrix';
identificationParam(1).Unit = '';
identificationParam(1).Value = [a11, 0, 0;
0, a22, a23;
0, a32, a33];
identificationParam(1).Minimum = [-inf, 0,0;
0, -inf, -inf;
0, -inf, -inf];
identificationParam(1).Maximum = [0, 0,0;
0, 0, inf;
0, inf, 0];
identificationParam(1).Fixed = [0, 1, 1;
1, 0, 0;
1, 0, 0];
end
Thank you,
mlut

4 Comments

This may actually be a bug, as my idnlgrey model reports a double '@' for the function handle used to create it.
model =
Continuous-time nonlinear grey-box model defined by '@@(t,x,u,p1,p2,varargin)simple3DOFmodel(x,u,p1,p2,actuationModelParamMat,boxyUSVmodelParam)' (function handle):
I have encountered the same error message. Any idea on how to fix it?
Hi Bernardo,
what I did is to enable "pause on errors", and then checked what the error was actually about. In my case it had nothing to do with the parameters. In my case I didn't expect to the 'fileArgument' of the ode function to be encapsulated in a cell array.
Biggest recommendation after trying to work with it for some time now: Don't use the toolbox, write your own code. There are a ton of bugs, the interface is horrible and the error reports not helpful.
Best regards,
Max
I cant figure out the fix it keeps bring me to line 59 and im getting the same error. I read through the documentation and the file arguement is expected to be in a cell array, but what did you do to actually fix it?

Sign in to comment.

Answers (0)

Products

Release

R2022b

Asked:

on 30 Jan 2023

Commented:

on 27 Sep 2023

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!