About units of params Accel Random Walk and Rate Random Walk

[narutojxl]

Hi author,
Thanks for your sharing so convient tool to community very much.
In code we average three axis Accel Random Walk to get accelerometer_random_walk of kalibr. accelerometer_random_walk's unit is m/s^3/sqrt(HZ), namely m/s^2/sqrt(s), but the Accel Random Walk's unit is m/s^2, is there a typo ? The same question exist about Rate Random Walk. Velocity Random Walk and Angle Random Walk units are equal to accelerometer_noise_density and gyroscope_noise_densityof kalibr. Thanks for your help and time!

allan_variance_ros/scripts/analysis.py

Line 121 in 25a3388

yaml_file.write("accelerometer_random_walk: " + repr(average_abi) + " \n")

[raabuchanan]

Hello,
There is actually a typo in the Kalibr documentation as discussed in this issue: ethz-asl/kalibr#354 and in this issue of GTSAM: borglab/gtsam#213

So accelerometer_random_walk should actually be m/s^2 and not m/s^2/sqrt(s),

I believe the other parameters have the same units as in the Kalibr documentation so if there is something not matching, could you be more specific? Thanks

[narutojxl]

Hi author, thanks for your help very very much.
According to kalibr gyroscope_random_walk unit is rad/s^2/sqrt(Hz), namely rad/s/sqrt(s). Your Rate Random Walk unit is rad/s, so i mistake there is also a typo in Rate Random Walk, sorry for that.
Maybe kalibr units of \sigma_{bg} and \sigma_{ba} should be corrected by someone. My problem is solved :)

[mintar]

Actually, the Kalibr documentation is correct, and ethz-asl/kalibr#354 and borglab/gtsam#213 are wrong. I've already replied in those two issues.

@narutojxl wrote:

accelerometer_random_walk's unit is m/s^3/sqrt(HZ), namely m/s^2/sqrt(s), but the Accel Random Walk's unit is m/s^2, is there a typo ? The same question exist about Rate Random Walk.

You were almost correct. The units should be m/s^3/sqrt(s) and deg/s^2/sqrt(s) (or equivalent). PR #4 fixes that.

[narutojxl]

Hi @mintar, thanks for your detailed explaination. I read in your PR4, and kalibr/issues/354 comment.
sigma_bg unit is (rad/s)/sqrt(s) or (deg/s)/sqrt(s), why it is deg/s^2/sqrt(s) ?
sigma_ba unit is (m/s^2)/sqrt(s), why it is m/s^3/sqrt(s)?
I'm a little confused about this and don't understand it . Could you please give me some info, thanks a lot.

You were almost correct. The units should be m/s^3/sqrt(s) and deg/s^2/sqrt(s) (or equivalent). PR #4 fixes that.

[mintar]

@narutojxl : Great catch! Yes, you were right all the time. The tables I've posted are correct, but I made a mistake in my comment to you and in #4. I've made a pull request to fix this (#8).

[narutojxl]

Thanks @mintar :).