I was just wondering if someone has already tried some sort of inertia navigation on arducopter platform? I at least could not find all that many discussions regarding this matter so I would assume there is something essential that I am missing.
Reasons that might hamper the inertia navigation (or atleast some stuff that popped to my mind):
-Noise on accelerometers. I have no idea how much aliasing there is from the vibrations etc. or if the accelerometers are nearly sensitive enough to be used for navigation. But I would at least assume that they might be able to get the location with accuracy of around 50cm (when drift corrected with gps every few seconds). But then again, I might be completely off on this one since I have no idea on performance of those accelerometers.
-Computing power. I would at least suppose that there should be enough power left to rotate the accelerometer values with rotation matrix and sum that up a couple of times to get location, and possibly to do some simple fusion with gps. but can this be too much for the arduino?
-No need. If there is just too much drift so that it cant be corrected with gps, or that the data would just make gps perform worse, then I would see that there is no use for inertial navigation. But other than that, I would see that inertial position would be extremely valuable data on loiter for example.
I was thinking of doing some simple testing with inertia at some point, but I just thought that it might be a good idea to ask here first if someone has already been tinkering with inertia, and possibly ruled that out completely for some reason.
So do you think that inertia navigation would be possible/useful on arducopter?
In the latest code I have commented out inertial alt hold. It's just too noisy. Try it and play around you may have better luck.
I had to rotate the Accel data with the DCM.