View Single Post
  #2   Spotlight this post!  
Unread 29-12-2014, 06:14
Michael Hill's Avatar
Michael Hill Michael Hill is online now
Registered User
FRC #3138 (Innovators Robotics)
Team Role: Mentor
 
Join Date: Jul 2004
Rookie Year: 2003
Location: Dayton, OH
Posts: 1,572
Michael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond reputeMichael Hill has a reputation beyond repute
Re: Question about the I term in the PID

According to conventional control theory, you shouldn't have to because, in general, you move, say, an arm up (which causes a positive integral term), then you'll move it down (offsetting back to 0). However, when dealing with real systems, a phenomenon known as integral windup occurs (http://en.wikipedia.org/wiki/Integral_windup). It's most noticeable when your motors go into saturation and just accumulate integral error. Theoretical motors don't have any saturation point, but real motors...well, you can feed in more gain, but once they're in saturation, they're not going any faster. In the wiki link, they outline a few ways to prevent the phenomenon. It's generally system dependent, so try them out and see what works best for your system.
Reply With Quote