Avoid Local Minima ngreyest

Hello to everyone,

I am writing this post because I would like to know if there are some ways to avoid local minima in parameter estimation done by nlgreyest. I've discovered the global optimization toolbox but it seems to do not work with idnlgrey data types, at least directly. Do you know if there are alternative ways to use it in grey box parameter indentification ?

Thanks in advance.

Answer 1

[Meet]

Hi Gianpio,

You can run "nlgreyest" multiple times using different initial parameter values to explore various local minima. Alternatively, you can set bounds on the parameters within the "idnlgrey" model to restrict the optimization search space.

Hope this helps!!