Minor oscillations around the desired position

Hi Everyone, Thanks to the package, I got my position controller to work using motion capture data from optitrak. I have a small issue with the precision and would like to get some tips on improving the accuracy.

Currently pelican is achieving an accuracy of +/- 5 cm around the desired point with motion capture frequency data of 30Hz. I would like to reduce this oscillation to +/- 2cm or even lower. Is there any way to stop these low frequency oscillations? I noticed that the quadcopter has a drift when I do not switch on the controller and am in acc mode. I do not know if this is the cause for the oscillations. Can any one suggest tips for improving the accuracy of the controller?

Thanks, Gowtham

With the luenberger observer, the acc biases may be a little unstable. You can try making them slower as described in the tutorial http://wiki.ros.org/asctec_hl_interface/Tutorials/hlp%20position%20control#Onboard_State_Estimation . Alternatively, the bias estimation in our state estimation frameworks is much better: http://wiki.ros.org/ethzasl_sensor_fusion

And actually just released: https://github.com/ethz-asl/ethzasl_msf/wiki

Also, in these frameworks, you can experiment with the noise parameter for the biases.

Cheers, Markus

From: garimellagowtham notifications@github.com<mailto:notifications@github.com> Reply-To: ethz-asl/asctec_mav_framework reply@reply.github.com<mailto:reply@reply.github.com> Date: Tuesday, October 29, 2013 2:06 To: ethz-asl/asctec_mav_framework asctec_mav_framework@noreply.github.com<mailto:asctec_mav_framework@noreply.github.com> Subject: [asctec_mav_framework] Minor oscillations around the desired position (#12)

Thanks, Gowtham

— Reply to this email directly or view it on GitHubhttps://github.com/ethz-asl/asctec_mav_framework/issues/12.

ethz-asl / asctec_mav_framework

Minor oscillations around the desired position #12