Sensor variance settings

[Arto]

Hi

How do sensor variance settings affect YAW drift? There is still drift in yaw while magnetometer update is not on and GPS with good lock is on.
I do not have magnetometer updates on but still default value for magnetometer variance is 1 in sensor variable settings. GPS variance is 10 and accelerometer is 25. Why is magnetometer variance 1 if the update is not on by default?
Can you hint how does changing these may affect the YAW. I beleive that by tuning these values it is possible to get rid of the yaw.

What are the best parameters for UAV plane ?

[CalebModerator]

Hi Arto,

The default settings work very well on small airplanes. The yaw angle is only observable when the sensor is accelerating. The longer it goes without acceleration, the more the yaw angle can drift.

[arto]

I put GPS variance from 10 to 1 and it really helped to remove the yaw drift. I tested this by my car. Today I will integrate GP9 into UAV and test how it works. Why did the changing the GPS variance from 10 to 1 helps so much to the drift when driving car? I was really speeding and accelerating with my Corvette

[CalebModerator]

Decreasing the GPS variance causes the GPS measurements to be assigned more weight in the filter, so that GPS correction happens more quickly. This causes the yaw angle to converge more quickly, and it causes the bias estimator to lock on faster. This would explain why the yaw drift decreases when you’ve decreased GPS variance.

The tradeoff is that any GPS error will cause larger yaw, pitch, and roll measurement errors.

[Author]

Posts