Can someone explain 2010 Rotate To Target VI

[willsterm]

Can someone provide a brief summary of how the Rotate To Target VI is supposed to work – in the context of the out-of-the box Update2 release of the 2010 FRC Framework – in particular in the context of Autonomous Independent mode? It’s error output when substantially small enough will force the Autonomous Independent to exit its while loop .. but I don’t understand why the Rotate To Target VI sets the PID’s setpoint in the fashion that it does. Can any one explain the rationale for this?

[Greg McKaskle]

The Rotate to Target subVI is provided with the target info. If there are zero identified targets, the setpoint will be initialized to zero degrees. If the target is found, it will be the absolute gyro position where the target is located.

When not initializing, the code uses the PID to determine the motor outputs and computes the error. When the error term is small enough, the robot is pointing at the setpoint, either zero degrees or the target. It leaves the loop because presumably you would like to kick, push, flail, or some other action. If there is an additional ball, you could then move to it, and execute similar code again, possibly by putting a loop around the code already there.

This is relatively simplistic, doesn't know how to trigger a kick, and assumes that when you are supposed to kick, it is better to kick to your end of the field (zero degrees), than to wait forever.

If you have additional questions about changing it, fire away.
Greg McKaskle

[willsterm]

I must be thick.

In autonomous is the first thing .. so presumably the gyro is set to zero (and the judges can re-position the robot as they see fit, so zero may not mean that the robot has a useful reference .. so if when the game starts the target is not in view the robot rotate to the zero angle .. which it currently believes it is in .. so no movement.

The other case is that the robot does see the target, so the setpoint is set the current heading .. which would be zero anyway ... once again the robot has not moved.


What am I missing?

Thx

[Greg McKaskle]

To the left of the loop there is a wait of zero, a reading of the global that extracts the number of targets. You may want to do something with this info before entering the loop, say wait for a few seconds or until you see a target, perhaps move forward to touch the ball. All of that seems like good stuff.

The code was written as a sample to demonstrate some of the game elements related to vision. It will not win any matches.

If there is a target visible fifteen degrees away from zero when auto begins, this code will rotate the robot to point at it. If you have questions as to how to get it to do something else, please ask.

Greg McKaskle

[willsterm]

Don't need help doing other neat stuff -- we are already doing neat thing in TeleOp mode -- but thanks for the offer.

What we are trying to is understand the underlying details of how and why the rotate to target works.

If we do what you say and put the find the target bit into the initial flattened sequence structure .. Then why do the next bit.

[Greg McKaskle]

You are welcome, but I meant for autonomous.

The underlying rotate to target VI expects to be called repeatedly at a relatively consistent interval when the robot needs to be rotated to aim at the target. If no target is currently visible, it uses zero for its setpoint. For each call, it reports back with the error between where it is being asked to aim and where it is currently aimed.

Each time the VI is called, it will update the motor PWM settings attempting to pivot the robot so that the target is in the center of the image. If the VI is being called in the loop and something pushes the robot, it will push back to try and get back to the angle. The VI calculates the values for the motors using a PID VI. The PID is tuned for a specific robot very similar to what was demoed at kickoff. The PID can be retuned for other types of robots. If you need assistance doing the tuning, there is a white paper on the NI site that covers this.

I don't understand the last question, so I hope this was more what you are looking for. If not, please ask specific questions.

Greg McKaskle

[willsterm]

This document has all the answers that we were seeking!!!

'Guide to the FRC 2010 Sample Source'

http://www.usfirst.org/roboticsprogr....aspx?id=10934

[Vikesrock]

Originally Posted by willsterm

...
and the judges can re-position the robot as they see fit
...

1) Judges judge awards, referees are the ones in charge of making calls during matches

2)The referees certainly cannot reposition your robot as they see fit. You may be thinking of <G11> but that only comes into play if you try to use an external alignment device to set up your robot (ruler, protractor, etc.)