PID Control Basics

PID Control Basics

PID Tuning

Rob Sink Technical Support Specialist

June 14th, 2016

Copyright © Yokogawa Corporation of America

1

What will be covered: 1.  2.  3.  4.  5. 

Common Process Control Techniques Process Dynamics What is PID PID Control Components How to Tune a PID Loop


2

Why do I Need to Understand PID

Ø  Every process is different Ø  Makes manual tuning easier Ø  Helps companies save money Ø  Helps facilities remain safe


3

Common Process Control Techniques ques Ø  Manual Control Ø  ON / OFF Control Ø  Closed Loop Control


4

Manual Control Ø  Operator observes the process error and adjusts the control output

Set Point

Δ

PID CONTROL

Measurement (Process Variable)

Process


5

ON / OFF Control Ø  Simplest form of feed back control Ø  Can be used for processes not requiring extremely tight control


6

Closed Loop Control Ø  The PID controller measures the process variable, compares it to the setpoint and then manipulates the output accordingly. Set Point

Δ PV

Measurement (Process Variable)

Final Control Element Copyright © Yokogawa Corporation of America

7

Process Dynamics: Dead Time Ø  Dead time is defined as the time before the process variable BEGINS to react to a change in the control output

Output

Process Variable Lag Time Dead Time


8

Process Dynamics: Lag Time Ø  Lag is defined as the time required for the process variable to adjust to a steady state after an output change is performed Ø  Lag time affects the control action

Output

Process Variable Lag Time


9

Process Dynamics: Output vs. Process Change


10

What is PID? Ø  PID control refers to process control using the coefficients Proportional, Integral and Derivative Ø  It is not P&ID which refers to Piping & Instrumentation Diagram


11

PID Control Defined Ø  PID control can be described as a set of rules with which a precise regulation of a closed-loop control system is obtained.


12

Temp (PV)

Temp Setpoint (SP)

PID Control Terms Ø  Proportional Band adjusts output amplitude (reciprocal of Gain) Ø  Integral eliminates offset error (automatic Reset or simply Reset) Ø  Derivative looks at the rate of change of the error (Rate)


13

Proportional Band Ø  The Proportional Band (P) is defined as the range over which the control output is adjusted from 0-100% Ø  Proportional does the heavy lifting getting the temperature close to the setpoint Ø  Some manufacturers use Gain instead of Proportional Band


14

Proportional with Manual Reset Ø  With proportional only control, an offset will be present between set point and process variable. Ø  Manual Reset allows a user to bias or shift the output to compensate for the steady state offset. 1000º Manual Reset Adjusted Here

Proportional Band

500º Set Point

Time Copyright © Yokogawa Corporation of America

15

Integral Ø  Integral action is used with proportional to eliminate the inherent offset Ø  The integrating term observes how long the error has existed, summing the error over time Ø  The sum becomes a value added to the output Integral Action Proportional Action

Output

+10%

Error -10%

Time

200 sec/repeat

Integral Time Constant


16

Integral – cont. Ø  Engineering units: Repeats/minute Minutes/repeat Seconds/repeat Ø  The integral action ceases at a no error condition


17

Integral at Work Setpoint

I

I

I

I

Integral started.

Each time period where the error is not zero, the output is increased (or decreased) by the Integral term. Copyright © Yokogawa Corporation of America

18

A Note About Integral Windup Ø  Integral windup refers to the situation in a PID controller where the integral, or reset action continues to integrate (ramp) indefinitely Ø  This usually occurs when the controller's output can no longer affect the controlled variable, which in turn can be caused by controller saturation Ø  Typical causes of Integral Windup are: The input has been removed from the process, output device has failed, a furnace door has been opened keeping the process from reaching temperature


19

Derivative Ø  Engineering units: minutes or seconds Ø  Anticipates the error rate and applies the “brakes” Ø  Derivative has no effect if the error is constant Derivative Action Integral Action

Output

+10%

Error -10%

Time

50 seconds

Derivative Time Constant


20

P, I and D Working Together

P only

P and I


21

PID

How to Tune a PID Loop Ø  Manually tuning the loop Ø  Using the controllers Auto/Self Tune


22

Manually Tuning a Loop Ø  These values are good starting points Ø  Change only (1) term at a time Ø  Make small changes observing the result


23

Fine-tuning the Proportional Band Ø  Work from larger to smaller numbers (wider to narrower) Ø  If cycling appears, the proportional band is too narrow


24

Fine Tuning the Integral Time Ø  The main goal is to reduce the offset Ø  Adjust from longer to shorter time Ø  If an oscillation exists at a longer period then the integral time is too short


25

Fine Tuning the Derivative Time Ø Adjust from shorter to longer time Ø If short-period oscillations develop, the time is to long. Ø The larger the Derivative, the stronger the corrective action and the more likely the output will become oscillatory


26

Tuning Loops with Dead Time Set P to 5% and the I & D to 0% Start the process with a setpoint that will allow the process variable to stabilize

Output

Dead Time

Lag Time


27

Process Variable

Tuning Fast Reacting Loops Set P to 100% and the I & D to 0% Start the process with a setpoint that will allow the process variable to stabilize


28

Using Auto Tune to Determine PID Values Ø  The output is varied between 0% and 100% three times (these values may be limited). Ø  The process variable must ascend and descend through set point for the output to change state. Ø  The auto tune algorithm observes the PV response to these output changes and installs the appropriate PID terms.


29

Ways to Prevent Overshoot

Ø  Limit the working output or enable an output ramp rate (if available) Ø  Limit the output range which will have an effect on the time it takes to get to setpoint Ø  Ramp the setpoint at a slow rate Ø  Use fuzzy logic (if available)


30

Fuzzy Logic Ø  Fuzzy logic is used to help reduce setpoint overshoot Ø  Used in addition to PID control


31

Yokogawa Products that Use PID Control

PLC

Single loop controller

Programmable controller

PLC/RTU

DCS Copyright © Yokogawa Corporation of America

32

UTAdvanced Line of controllers Ø  1-2 loops of control Ø  Built in ladder sequence control Ø  Software used in Webinar Ø  Nuclear qualified


33

YS1000 Family of Controllers

Ø  1-2 loops of control Ø  Nuclear qualified Ø  Hard manual backup Ø  Function block programming


34

FA-M3 PLC

Ø  Modular PLC design Ø  4 control loops per PID module Ø  PID control is not done in ladder logic


35

Questions

Questions? Feel free to email us with further questions at [email protected]. Please put “PID Webinar” in the subject line.


36

Thank you for attending! Feel free to email us with further questions at [email protected]. Please put “PID Webinar” in the subject line.


37

PID Control Basics

Recommend Documents