Visual servoing #545
Visual servoing #545
Conversation
| @@ -82,14 +83,14 @@ public void initialize() { | |||
| ySetpoint = m_swerve.getState().y(); | |||
| thetaSetpoint = m_swerve.getState().theta(); | |||
|
|
|||
| Optional<Pose2d> opt = m_fieldRelativeGoal.get(); | |||
| Optional<Pose2d> opt = m_fieldRelativeGoalSupplier.get(); | |||
There was a problem hiding this comment.
it looks to me like this command only does the "coordinated profile" thing if it happens to run with a goal the first time it is "initialized". also, note that the coordination scaling changes the profiles, so if you press the button over and over, the profiles will get slower and slower. maybe take out all this initialization coordination stuff until the basic uncoordinated profile is working perfectly?
There was a problem hiding this comment.
I figured the whole calculation might take too long to run each cycle, however I could be wrong about that, so i'll change that, the profile does not get slower and slower due to the fact the scaling is undone in the end function.
There was a problem hiding this comment.
tbh I would suggest leaving the coordination part out until the rest works perfectly.
There was a problem hiding this comment.
if you want to keep it, definitely don't do the "undo at the end" thing. just don't overwrite the original.
| Rotation2d bearing = new Rotation2d( | ||
| Math100.getMinDistance(measurement, fieldRelativeGoal.getRotation().getRadians())); | ||
| Math100.getMinDistance(measurement, m_fieldRelativeGoal.getRotation().getRadians())); |
There was a problem hiding this comment.
note the word "bearing" means "angle from point A to point B" but the expression here appears to return simply the angular component of the goal pose itself, which isn't the same thing. what do you mean this to mean?
| @@ -159,7 +161,7 @@ public void execute() { | |||
|
|
|||
| @Override | |||
| public boolean isFinished() { | |||
| if (!m_fieldRelativeGoal.get().isPresent()) return true; | |||
| if (!m_fieldRelativeGoalSupplier.get().isPresent() && m_fieldRelativeGoal == null) return true; | |||
There was a problem hiding this comment.
don't you want the command to keep going even when it loses sight of the target?
There was a problem hiding this comment.
See other comment about null.
| if (opt.isEmpty()) { | ||
| if (m_fieldRelativeGoal == null) |
There was a problem hiding this comment.
this logic seems to make the command end one cycle after the target is lost.
There was a problem hiding this comment.
This part is mainly to prevent the case where when I press the button no note is identified having some random result. This will only happen if no object has ever been detected since the robot has been booted due to the fact it only is null upon initialization of the class and is changed upon the first initialization of the command given a pose is given.
There was a problem hiding this comment.
hm, I think it's not doing what you think it's doing. maybe start with something simple that can work perfectly.
| @@ -55,7 +56,7 @@ public DriveWithProfile2( | |||
| SwerveDriveSubsystem drivetrain, | |||
| HolonomicFieldRelativeController controller, | |||
| SwerveKinodynamics limits) { | |||
| m_fieldRelativeGoal = fieldRelativeGoal; | |||
| m_fieldRelativeGoalSupplier = fieldRelativeGoal; | |||
| m_swerve = drivetrain; | |||
| m_controller = controller; | |||
| m_limits = limits; | |||
There was a problem hiding this comment.
these profile limits are unattainably fast, aren't they? did you try making the profiles a lot slower, as i suggested? "first make it perfect, then make it fast"?
No description provided.