Gyro math filters

[ekapalka]

Originally Posted by SoftwareBug2.0

Do you want to do a full 6 degree of freedom system or do you want to assume that you stay on flat ground?

I'm assuming that the robot is always on a flat surface

Originally Posted by mathmogul

It isn't possible to use an accelerometer for measuring yaw rotation.

That's pretty much the conclusion I arrived at. There must be some correlation between rotating and accelerometer output, though, like centripetal acceleration?

A magnetometer should be used instead since it's essentially a compass.

I've heard mixed opinions about using compasses. We already have one of these, so I guess we'll modify the ADXL345_I2C class into a magnetometer class (using its pre-written i2c code) and see how it goes from there