Register to reply

Fuzzy logic control vs PID?

by I_am_learning
Tags: control, fuzzy, logic
Share this thread:
I_am_learning
#1
Jan28-13, 10:31 PM
I_am_learning's Avatar
P: 669
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.
Phys.Org News Partner Engineering news on Phys.org
UConn makes 3-D copies of antique instrument parts
Amazon worker piloted drone around Space Needle
Five next-generation technologies for positioning, navigation and timing
Mordred
#2
Jan28-13, 10:50 PM
Mordred's Avatar
P: 1,857
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.
AlephZero
#3
Jan29-13, 09:56 AM
Engineering
Sci Advisor
HW Helper
Thanks
P: 6,948
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.

I_am_learning
#4
Jan29-13, 10:12 AM
I_am_learning's Avatar
P: 669
Fuzzy logic control vs PID?

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?
Mordred
#5
Jan29-13, 05:11 PM
Mordred's Avatar
P: 1,857
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...%20Control.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.
I_am_learning
#6
Jan29-13, 10:27 PM
I_am_learning's Avatar
P: 669
Quote Quote by Mordred View Post
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...%20Control.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.
Thanks.
The article looks just like what I was looking for.
I will read it thoroughly and come with more questions if required.


Register to reply

Related Discussions
Is 'Fuzzy' logic an appropiate Quantum logic ? Quantum Physics 2
What is fuzzy logic? Set Theory, Logic, Probability, Statistics 15
Fuzzy Logic General Discussion 10
Fuzzy Logic General Discussion 84