bode plot from experiment data

hello all, I have experimental input and output data (saved as data in excel).

- Data is sinusoidal function(time domain). The input signal have a continuously increasing frequency(ex. 0 Hz ~ 1000 Hz). And output signal is sensor measuring data that it have same shape as the input signal.

- I don't have transfer function.

I want draw bode plot using these experiment data. please some body help me.

Answer 1

[Nick Choi]

Assuming you have the input data and output data in the MATLAB workspace as vectors, you could use the 'tfestimate' function to estimate the transfer function of the data and then use the 'bode' function to obtain the bode plot of the transfer function.

The following documentation page provides additional information on how to use the 'tfestimate' function:

http://www.mathworks.com/help/signal/ref/tfestimate.html

If the data is not in MATLAB yet, you can use the 'xlsread' function to import the data from the Excel spreadsheet. The following documentation page provides additional information on how to use the 'xlsread' function:

http://www.mathworks.com/help/matlab/ref/xlsread.html

Update by Kishen Mahadevan at MathWorks on 6/19/2024

To get a dynamic system model ’sys’ for use with the ‘bode’ function, you can use the ‘tfest’ function from System Identification Toolbox to estimate the transfer function model.

Please refer to this documentation page for more information on ‘tfest’ - Estimate transfer function model - MATLAB tfest (mathworks.com)

Additionally with System Identification Toolbox you can also automatically estimate different dynamic models such as ARX and state-space models which can then be used with the 'bode' function to plot the bode diagram using ‘bode(sys)’.

[Alfred]

Does this still work on 2017b? I checked the tfestimate help page but the output of tfestimate is not a sys object, so I can't use bode(sys)