Perhaps he means this: Ax, Ay, Az, Vx, Vy, Vz, X, Y, Z.
i.e., 3 axes each for acceleration, velocity, and position. One integration gives velocity. Two integrations gives position. However, finite errors in acceleration measurement, if not predetermined and nulled, will also be integrated twice to produce terrible drift errors in the position values.
@strange3141 That is true, do you have any idea how much error such a system typically accumulates over time? the INS entry on wiki equates to about 7.1 inches per second of IMU Drift, is the drift really so high even a kalman filter implemented?
3: X,Y,Z (positions),
+3: (angle or angle-rate) of twist around X, Y, & Z,
That's the usual 6 degrees of freedom;
+3: guessing: (magnetic field) with portions in directions X, Y, Z.
{px,py,pz, rx,ry,rz, mx,my,mz} is a
very common basic sensor-set in satellites.
prfesser56 2 years ago
@prfesser56
yep, thats a typical 9DOF. U also find this configuration in avionics (quadrocopters, ..).
Maverickx89 1 year ago
@prfesser56
rezymartin 1 year ago
9-axis? come again?
where the hell did you find that many axes?
i only have 3 :(
solidacid1337 2 years ago
Perhaps he means this: Ax, Ay, Az, Vx, Vy, Vz, X, Y, Z.
i.e., 3 axes each for acceleration, velocity, and position. One integration gives velocity. Two integrations gives position. However, finite errors in acceleration measurement, if not predetermined and nulled, will also be integrated twice to produce terrible drift errors in the position values.
strange3141 2 years ago
@strange3141 That is true, do you have any idea how much error such a system typically accumulates over time? the INS entry on wiki equates to about 7.1 inches per second of IMU Drift, is the drift really so high even a kalman filter implemented?
MHUnited1 1 year ago