Questions about extracting White Noise and Flicker Noise (or IRBS)

[GeoNumericsMarc]

Hi!

My name is Marc Fernández, and I am currently working on a project that involves extracting White Noise and Flicker Noise (IRBS) from IMU data using the "allantools" library in Python. I would appreciate your expertise and guidance on a couple of issues I am encountering.

In my Python script, I load an IMU file contaminated with White Noise and IRBS, run the "allantools" function, and generate a plot of Allan variance. The challenge I face is accurately extracting the White Noise and IRBS from this plot.

To extract White Noise, I measure the intersection of the line when tau=1. For IRBS, I look for the point where the slope is zero and then divide by 0.664 (since IRBS = 0.664B).

For instance, in the attached graph, the data corresponds to an IMU with a 500Hz sampling rate over 4 hours, contaminated with White Noise of 201 º/h (0.001 rad/s) and an IRBS of 13 º/h (0.000063 rad/s). From the graph:

White Noise at tau=1 gives 3.15725E-05 rad/s.
IRBS, where the slope is zero, gives IRBS = 1.43653E-05 rad/s, which, when divided by 0.664, results in B = 2.16345E-05 rad/s.
Could you kindly review the attached Python code and provide any insights on whether it is correctly implemented? Additionally, could you advise on the best method to accurately extract White Noise and Flicker Noise (IRBS) from the plots?

Thank you very much for your time and assistance.

Best regards,

Marc Fernández Casanova