View Single Post
  #16   Spotlight this post!  
Unread 28-10-2013, 21:41
MichaelBick MichaelBick is offline
Registered User
FRC #1836 (MilkenKnights)
Team Role: Alumni
 
Join Date: Oct 2011
Rookie Year: 2010
Location: Los Angeles
Posts: 734
MichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant futureMichaelBick has a brilliant future
Re: Optical sensors getting tricked

Quote:
Originally Posted by yash101 View Post
Programming a consistent robot closed loop can be hard because you need to calculate the motor speed using the hundreds of variables
Not only is PID not required, you do not need hundreds of variables. A bang-bang shooter algorithm is incredibly easy to program(it's an if statement, if your programmers can program the robot to shoot they can program bang bang). Furthermore it only needs an encoder/ir sensor(we used banner sensors) to feed it RPM. As long as you know the speed of the wheel you don't need to account for any other variables.

For position PID is great. Though we haven't used it before, I have heard that limiting the controller to PD or just P is still a pretty darn effective motion controller.

To sum up, sensors don't have to be complicated. Even with complex code(like vision code) teams like 341 have made it much simpler and easier for the rest of us.
__________________
Team 1836 - The Milken Knights
2013 LA Regional Champions with 1717 and 973
2012 LA Regional Finalists with 294 and 973
To follow Team 1836 on Facebook, go to http://www.facebook.com/MilkenKnights
To go to our website, go to http://milkenknights.com/index.html

Last edited by MichaelBick : 28-10-2013 at 21:47.