Hi ,
Hi have a question, Is possible fusion all sensor IMU + GPS + Barometer with kalman filter to estimate and compensate drift of GPS for have a precise altitude , precise velocity, and precise position of multicopter ? Are there a algorithms o code?
Replies
Yes, it is possible, and it would be really beneficial. I have made some tests for some simple position fusion with IMU and it makes the position much better and theoretically even faster. But the thing is that full blown Kalman filter wont be able to run at 100Hz on the arduino, it would be more like 1Hz etc. since the floating point capabilities of the arduino are somewhat lacking.
Now many of the limitations can be solved, one possibility is to just cheat and not do a full Kalman update on each step, but just propagate the position and attitude according to gyros and accelerometers (Kalman prediction), and then correct for the drift with Kalman update step at 1Hz etc. But as far as I know, this is not quite sound solution, and there might be some extra complications on doing such things.
is it possible? probably.
Has it been done yet? Nope.
Do you have experience programming kalman filters? I've often wondered if they would be beneficial with respect to sensor fusion and redundancy.