Ok. The probability of anything happening is small but security first, right? So what I thought of is that has anyone tried to program an INS based on these cheaper IMU-units, like UAVDevBoard? The problem isn't programming but rather the accuracy of the sensors.
Besides security the INS could be used to reduce the error and lag in position between GPS updates which would make it possible to write navigation algorithms with much higher coefficients and, thus, tighter control.
Thoughts?
You need to be a member of diydrones to add comments!
Replies
We (my neighbor and myself) are working on a tightly-coupled GPS/INS implementation using a UBlox GPS and the CHR6d. We started with this IMU from Sparkfun and just coded an INS just to see how bad it would perform. We were getting pitch and roll errors of > 10 degrees after about 20 seconds of gentle rotations. Fast movements and rotations made it even worse.
This is why people here are implementing an AHRS instead of an INS. An AHRS has essentially zero drift when in steady-state and has some sort of heading/course information.
Regardless, ArduPilot Mega has pins for you to connect an optional external (to avoid interference) magnetometer if you want to use it in a GPS-denied environment or just have the extra data for safety.