1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Fuzzy logic control vs PID?

  1. Jan 28, 2013 #1
    I am trying to begin with fuzzy logic, but this initial question is preventing me from moving any forward.
    It appears that fuzzy logic control is just an improvement over bang-bang control system. Like instead of doing bang-up , bang-down it does bang-up-hard, bang-up-small bang-down-small etc. It just appears to be discretized proportional control. If it is all about it, we have even better PID control, so why use fuzzy? Obviously I am missing something. Please help.
  2. jcsd
  3. Jan 28, 2013 #2
    What your missing is the needed process model. In siuations where close is good enough the bang up bang down method is fine. One example is tank filling.

    However there are numerous situations that require tighter control. Some will accept a quartlery amplitude decay. Some need even further dampening to being critically dampened.

    The key is the needs of the process.
  4. Jan 29, 2013 #3


    User Avatar
    Science Advisor
    Homework Helper

    Most "simple" control theory (e.g. PID system design) assumes the behaviour of the system is lnear and its transfer functions are known.

    Fuzzy logic system design doesn't require either of those two assumptions.
  5. Jan 29, 2013 #4
    In fuzzy logic control of temperature, you don't say temp is cold, instead you say temp is .83cold, so instead of doing heater on you do heater .83on.
    I still find it fundamentally very similar to proportional control.
    Can someone show me fundamental difference?
  6. Jan 29, 2013 #5
    Rather than go into th extensive pros and cons of each this article has a well described comparison between the two. In many ways either can be used.

    http://www.ece.uidaho.edu/ee/classes/ECE573F05/PID vs Fuzzy Control.pdf

    one thing to note that this article does not discuss ifs the Kreiger Nickoles math usually give a quarterly amplitude decay.

    thats 4 overshoots each overshoot less each time before it stabilizes on a setpoint. The math gets you close but not critically dampened. critically dampened is no overshoot a smooth transition to set point.
    Last edited: Jan 29, 2013
  7. Jan 29, 2013 #6
    The article looks just like what I was looking for.
    I will read it thoroughly and come with more questions if required.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook