- #1
PID algorithm for constant temperature controller.
- Thread starter gputowski
- Start date
Click For Summary
Discussion Overview
The discussion revolves around developing a PID algorithm for a heating controller, specifically addressing the regulation of relay timing for heating cycles. Participants explore the nuances of PID control in the context of time lags and the relationship between ON and OFF states of the relay.
Discussion Character
- Technical explanation
- Debate/contested
- Mathematical reasoning
Main Points Raised
- One participant seeks assistance in writing a correct PID algorithm for a heating controller and mentions a graph for reference.
- Another participant questions the meaning of "correct PID algorithm" and provides the general formula for PID control.
- A participant highlights the importance of considering time lag in the heating regulation process, noting that it can be significant.
- Discussion includes the idea that the slope of the temperature curve during the initial heating cycle may indicate the process's thermal capacity and time lag.
- The original poster expresses familiarity with the PID formula but questions whether a single formula can regulate both ON and OFF relay times or if separate formulas are needed.
- Another participant points out a potential misunderstanding regarding ON and OFF times, suggesting that the relationship between them is not straightforward and introduces the concept of "Bang-Bang control." They also provide references for further reading.
- There is a suggestion to keep ON-OFF intervals short relative to process time constants and to consider using PWM at the controller output.
Areas of Agreement / Disagreement
Participants do not appear to reach a consensus on the best approach to regulate the heating controller. There are multiple competing views regarding the handling of ON and OFF times and the application of PID versus Bang-Bang control.
Contextual Notes
Participants express uncertainty about the relationship between ON and OFF timing in the context of PID control and whether separate formulas are necessary. The discussion also reflects varying levels of understanding regarding control strategies and their implications.
- #2
gputowski
- 3
- 0
Any idea?
- #3
jim hardy
Science Advisor
Homework Helper
Dearly Missed
- 9,864
- 4,896
Just what do you mean by
One PID algorithm is
Controller output = proportional gain P X error + integral gain I X ∫(error) + derivative gain D X d(error)/dt
?gputowski said:
One PID algorithm is
Controller output = proportional gain P X error + integral gain I X ∫(error) + derivative gain D X d(error)/dt
- #4
Svein
Science Advisor
- 2,316
- 813
The one thing you need to remember when trying to regulate a heating controller is the time lag through whatever you are trying to regulate. This lag can easily run into hours.
- #5
jim hardy
Science Advisor
Homework Helper
Dearly Missed
- 9,864
- 4,896
Slope of temperature curve during that first on-cycle should give a clue about process thermal capacity, and maybe that time lag mentioned by svein..
- #6
gputowski
- 3
- 0
Thank You for the answer.
I know the formula for PID. But If You have seen my first drawing You should see that We are regulating the time when the relay is ON. We all supposed to regulate the time when the relay is OFF. Is it possible to make it by one formula or I will have to make two formula with time: first when the relay is ON second when the relay is OFF and make a loop.Greg
I know the formula for PID. But If You have seen my first drawing You should see that We are regulating the time when the relay is ON. We all supposed to regulate the time when the relay is OFF. Is it possible to make it by one formula or I will have to make two formula with time: first when the relay is ON second when the relay is OFF and make a loop.Greg
Attachments
- #7
jim hardy
Science Advisor
Homework Helper
Dearly Missed
- 9,864
- 4,896
I'm not sure i understand that statement.gputowski said:
ONtime is (1-OFFtime).
You're mixing linear control with" on-off control" here, sometimes that's called "Bang-Bang control"
and the math of that is another world. I won't go there myself
From a quick search:
Here's a paper...
http://dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica4.pdf
and a couple references
http://en.wikipedia.org/wiki/Bang–bang_control
http://cds.linear.com/docs/en/datasheet/1041fa.pdf
http://liberzon.csl.illinois.edu/teaching/cvoc/node86.html
Your best bet is to keep your on-off intervals short compared to the process time constants and treat it as linear
or insert a PWM at your controller output.
Similar threads
- · Replies 14 ·
- · Replies 12 ·
- · Replies 10 ·
- · Replies 6 ·
- · Replies 1 ·
- · Replies 5 ·