How to set prediction horizon explicitly in pem() -function for parameter identification?

Hi,

I'd like to test pem() -function with different N-step long prediction horizons such that the parameter identification would be done based on N-step prediction (multiple step prediction) with the given model. Where can I adjust the prediction horizon with this function when called in command line?

Answer 1

[Adithya Addanki]

Hi Joonas,

Please refer to the 'opt' argument section in the link below:

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

You may be able to do this using the first model type (idss). For more information please refer to the usage of the same.

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

Thank you, Adithya

[Joonas]

Hi Adithya,

unfortunately I'm using idgrey model as a model type and I'd like to adjust the prediction horizon in pem() -function with this grey-box model. Is this possible? I haven't found a way to do this so far.

Thanks anyway for your suggestion! I think that it applies only for idss -models.

Any solutions with idgrey models?