When filtering data using the system identification app, does it account for the filter phase lag?

im using system identification app to creat transfer function. my data is noisy, i want to filter the data between 0 and 20hz does the system identification will take in a account the lag added by the filter i chose?

thank you.

Answer 1

[Ayush]

Hey Eyal,

I understand that you want to filter to the data between 0 and 20Hz before performing system identification to remove noise from it. You want to know whether the system identification app considers the time delay introduced by the filter or not.

Yes, the System Identification Toolbox in MATLAB does have the capability to handle delays introduced by filters. The toolbox provides various functions and algorithms specifically designed for system identification, including methods for dealing with time delays.

When using the System Identification Toolbox, you can specify the presence of a delay in your system model. This can be done by using the “idtf” function to create transfer. This allows you to explicitly define the time delay parameter in your model.

Additionally, the toolbox provides advanced algorithms like the “delayest” function, which can automatically estimate the time delay present in your system based on the input-output data. This function can be used to estimate the delay and incorporate it into your system model.

Link for the above mentioned functions:

1. Transfer function model with identifiable parameters - MATLAB - MathWorks India
2. Estimate time delay (dead time) from data - MATLAB delayest - MathWorks India

Hope this helps!!

Regards,

Ayush Goyal