Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Algorithm for filter and AutoTuning Temperature Control

Status
Not open for further replies.

ServoMotionControl

Electrical
May 15, 2009
1
I am new in PLC domain. I am trying to write a PLC program for autotuning a PID for temperature control.
I am looking for algorithm for filtering the noise in the analog signal and a robust algorithm for autotuning. which method is recommended ?
Many thanks for your help.
Ad
 
Replies continue below

Recommended for you

Writing a good auto tune program in a PLC isn't practical. Most PLCs don't have the memory space or can easily manipulate matrices so the brute force method can't be used. The temperature controllers probably have some algorithms in firmware written in C that make good guesses using heuristics rather than a minimizing routine. If you can estimated the plant gain, time constant and dead time by using some heuristic method then you can use the formulas on to calculate the gains. I would read the material on the control guru site.

Have you written an auto tuning program for servo motion control or is this your first one?




Delta Computer Systems
 
PNachtwey hit the nail headwise. Even if the PLC in question is capable it isn't worth the effort to roll-your-own.

Certain PLC models have autotuning built in, but you'll likely end up hand-tuning the loop. Don't despair, because there are few better ways to develop an understanding of loop control other than the laying on of hands, and most temperature loops are amenable to hand-tuning.

One (potentially) useful cheat is to obtain a single loop controller with autotuning capabilities, temporarily wire it in, tune the loop, record the resulting values, remove it (and wire up the PLC), and use the controller's autotuned values as a starting point.

*Potentially* useful, because each PLC manufacturer seems to have a slightly different take on how to define and scale their P, I, D (and other) terms. I've seen everything from engineering units for I and D (minutes, or seconds) to ... well, on some I never have been able to figure out what the manufacturer had in mind. [nosmiley]
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor