I am using PID controller to stabilize quadcopter. Its working well on the jig. But when I removed jig it went up and hit ceiling. I analyzed my data and from which I come to the conclusion that at static poisition my gyro sensor is outputing +-6deg/sec but when I start motors (without being control in) gyro rate jupms to +-30deg/sec. This increase in rate due to vibrational noise is causing quadcopter to liftup without beign my intension. Any suggestions to get rid from this vibrational noise?
5
-
$\begingroup$ What axis is the gyro sensor measuring? $\endgroup$ – Ian Nov 11 '14 at 22:07
-
$\begingroup$ roll and pitch axis $\endgroup$ – arshad mahmood Nov 13 '14 at 10:38
-
$\begingroup$ What are you using to control desired/actual altitude? $\endgroup$ – Ian Nov 13 '14 at 18:01
-
$\begingroup$ I am using PID control $\endgroup$ – arshad mahmood Nov 20 '14 at 4:25
-
$\begingroup$ Sorry... I meant what sensor are you using to measure altitude so that you can control it. $\endgroup$ – Ian Nov 20 '14 at 16:17
Add a comment
|
$\begingroup$
$\endgroup$
2
You could probably use a low pass filter to eliminate the vibrational noise.
You should also correct the drift in your gyro measurement using some form of altitude sensor.
-
$\begingroup$ when I used low passfilter, it actually attenuating actual signal and i am loosing control. Noise is remove but actually low pass filter delays the output to one previous step. I am correcting that drift using accelerometer. $\endgroup$ – arshad mahmood Nov 11 '14 at 4:13
-
$\begingroup$ Low pass filtering is not working here. $\endgroup$ – arshad mahmood Nov 11 '14 at 4:13
$\begingroup$
$\endgroup$
1
I recommend you to use a kalman filter if you want to remove the noise without disfiguring the usefull part of the gyroscope's signal.
-
$\begingroup$ Even though the actual algorithm of Kalman Filter is simple, meeting the requirements of the filter is not easy besides the model must be specified accurately. $\endgroup$ – CroCo Jan 12 '15 at 11:22