If I want get angles roll, pitch and yaw relative to the orbital coordinate system (OCS), which is set as follows:
X_OCS - is directed along the velocity vector of the spacecraft,
Z_OCS - is in the direction of the local vertical (nadir)
Y_OCS - is complementary to the right-hand coordinate system,
and I want to get the angles relative to this coordinate system OCS, and not relative to the TOD, J2000 and other - how do I correctly write a code? frameS in J2000 and I used LVLH_CCSDS.
I used code:
Frame frameS = s.getFrame();
final GeodeticPoint targetDef = new GeodeticPoint(FastMath.toRadians(5.), FastMath.toRadians(-40.), 0.);
final TargetPointing targetLaw = new TargetPointing(orbit.getFrame(), targetDef, earthSpheric);
// Get roll, pitch, yaw angles corresponding to this pointing law
final LofOffset lofAlignedLaw = new LofOffset(orbit.getFrame(), LOFType.LVLH_CCSDS);
final Rotation lofAlignedRot = lofAlignedLaw.getAttitude(orbit, date, frameS).getRotation();
final Attitude targetAttitude = targetLaw.getAttitude(orbit, date, frameS);
final Rotation rollPitchYaw = targetAttitude.getRotation().compose(lofAlignedRot.revert(), RotationConvention.VECTOR_OPERATOR).revert();
final double[] angles = rollPitchYaw.getAngles(RotationOrder.ZYX, RotationConvention.VECTOR_OPERATOR);
for angular velocity and angular acceleration (rad/a, rad/s^2) - and for angles around Y and Z - angular velocity and angular acceleration correct, but for X - for example, for big angle around X - I get little (~10e-05) angular velocity - and it not correct and not equal real value.
But for Y and Z all good - the values around Y and Z are equal to the real values.
Why can this be? Maybe because something is set incorrectly, maybe the opposite direction is needed for X?
In this question, the sequence of angles is not important to me - I can set it as I need for different spacecraft. That’s why I just use Z, Y, X in the question - it doesn’t affect my questions.
First of all, beware that in the general case, you can’t have both X exactly along velocity and Z exactly along nadir, because velocity and nadir are not always exactly orthogonal to each other.
Local orbital frames are often either aligned exactly with velocity or aligned exactly with either position or opposite of position (which also is not nadir because it doesn’t account for Earth oblateness).
If you choose LVLH_CCSDS as your local orbital frame, you will get Z aligned with position. If you rather need X aligned with velocity, then you should use VNC as the local orbital frame.
I think the way you compute the angles is correct, but you really have to check signs and orders of rotations applications by yourself on simple cases, it is really easy to get them wrong. I suggest you set up test cases (perhaps even only on a debugger) with axes exactly aligned to canonical axes (say position exactly along inertial X axis and velocity exactly along inertial Y axis) and check with a few targets at simple places (just below satellite, just above satellite, just East of the satellite, just West of the satellite, just North of the satellite, just South of the satellite) and see if the expected angles are correct (I guess in these cases, only one angle should be non-zero). Once you have checked the angles signs are correct, then you can set up more complex test cases with intermediate points and look if the angles ordering is correct.
I have no questions about the angles and their values - they are equal to the real values.
That is, I set everything up correctly. However, I would like to know the code for finding these angles relative only to the orbital coordinate system (the center at the spacecraft’s center of mass) and relative to the inertial coordinate system.
Do I understand correctly that these are Euler angles, that is, at the first rotation, the second rotation is performed around the modified axis as in the picture?
@luc I have already done test calculations, for example, aiming at the nadir and obtaining almost zero angles of pitch, roll and yaw. Aiming strictly to the right and left - where I got approximately the roll angles calculated on paper with a minus and a plus. I also got a match with real data from a real spacecraft.
The only thing I would like to hear from You - is the search for angles relative to the orbital coordinate system and relative to the inertial coordinate system.
Also I need to figure out how to refine my angular velocity on the X axis.
- is the search for angles relative to the orbital coordinate system and relative to the inertial coordinate system.