ISA GOOD TUN PKT GUD-2015 Good Tuning A Pocket Guide (4th Edition).pdf

1、Good Tuning: A Pocket GuideFourth EditionGood Tuning:A Pocket GuideFourth EditionBy Gregory K. McMillanNoticeThe information presented in this publication is for the general education of the reader. Because neither the author nor the publisher has any control over the use of the information by the r

2、eader, both the author and the publisher disclaim any and all liability of any kind arising out of such use. The reader is expected to exercise sound professional judgment in using any of the information presented in a particular application. Additionally, neither the author nor the publisher has in

3、vestigated or considered the effect of any patents on the ability of the reader to use any of the information in a particular application. The reader is responsible for reviewing any possible patents that may affect any particular use of the information presented. Any references to commercial produc

4、ts in the work are cited as examples only. Neither the author nor the publisher endorses any referenced commercial product. Any trademarks or tradenames referenced belong to the respective owner of the mark or name. Neither the author nor the publisher makes any representation regarding the availabi

5、lity of any referenced commercial product at any time. The manufacturers instructions on use of any commercial product must be followed at all times, even if in conflict with the information in this publication.Copyright 2015 International Society of AutomationAll rights reserved. Printed in the Uni

6、ted States of America.1098765432ISBN 978-1-941546-43-7No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher.ISA67 Alexander

7、DriveP.O. Box 12277Research Triangle Park, NC 27709Library of Congress Cataloging-in-Publication Data in process.ISA wishes to acknowledge the cooperation of those manufacturers, suppliers, and publishers who granted permission to reproduce material herein. The Society regrets any omission of credit

8、 that may have occurred and will make such corrections in future editions.viiAcknowledgmentsWhen I retired from Monsanto-Solutia, I was fortunate to have the opportunity to contract as a consultant to Emerson Process Management at the headquarters in Austin in 2004. The support by Mark Nixon, manage

9、r of DeltaV R the diversity of processes, con-trol valves, control algorithms, and objectives makes this impractical.Myth 2 Controller tuning settings can be computed precisely. Not so. The variability and nonlinearity in nearly all processes and control valves makes this implausible. Any effort to

10、get much more than one significant digit is ques-tionable because any match to the plant is momentary. If you run a test for an auto tuner or manually compute settings ten times, you should expect ten different answers.2 BEST OF THE BASICSPulp and paper plant studies found 75 percent of the loops ca

11、used more variability in the auto-matic mode than in the manual mode. A third of them oscillated as a result of nonlinearities such as valve dead band. Another third oscil-lated because of poor controller tuning. The remaining loops oscillated because of deficien-cies in the control strategy. A well

12、-designed con-trol loop with proper tuning and a responsive control valve can minimize this variability. Because this means you can operate closer to constraints, good tuning can translate into increased production and profitability. 1.2 Actions Speak Louder than WordsThe very first settings that mu

13、st be right are the controller and valve actions. If these actions are not right, nothing else matters. The controller output will run off scale in the wrong direction regardless of the tuning settings.The controller action sets the direction of a change in controller output from its propor-tional m

14、ode for every change in the controllers process variable (feedback measurement). If you choose direct action, an increase in process vari-able (PV) measurement will cause an increase in BEST OF THE BASICS 3controller output that is proportional to its gain setting. Since the controller action must b

15、e the opposite of process action to provide feedback correction, you should use a direct-acting con-troller for a reverse-acting process except as noted later in this guide. Correspondingly, you should select reverse control action for a direct-acting process so an increase in process variable measu

16、rement will cause a decrease in controller output that is proportional to its gain setting, except as noted later. A direct-acting process is one in which the direction of the change in the process variable is the same as the direction of the change in the manipulated variable. A reverse-acting proc

17、ess is one in which the direc-tion of the change in the process variable is opposite the direction of the change in the manipulated variable. The manipulated variable is most frequently the flow through a control valve, but it can also be the set point of a slave loop for a cascade control system or

18、 variable speed drive.The valve action sets the display. For example, it determines whether a 100 percent output signal corresponds to a wide open or a fully closed valve. It also determines the direction of a change in the actual signal to the control valve 4 BEST OF THE BASICSwhen there is a chang

19、e in the controllers output. In some analog controllers developed in the 1970s, such as the Fisher AC2, the valve action affected only the display, not the actual signal. To compensate for this lack of signal reversal for a reverse-acting valve (i.e., an increase-to-close or fail-open valve), the co

20、ntrol action had to be the opposite of the action that would normally be appropriate based on process action alone. For-tunately, the valve action corrects both the dis-play and the actual valve signal in modern controllers, so the control action can be based solely on process action. However, the u

21、ser should verify this before commissioning any loops. In control systems that use fieldbus blocks, the valve action should be set in the ana-log output (AO) block rather than in the PID controller block. This ensures that the “back-cal-culate” feature is operational for any function blocks (split r

22、ange, characterization, and signal selection) that are connected between the PID and AO blocks. The signal can also be reversed in the current-to-pneumatic transducer (I/P) or in the positioner for a control valve. Before the advent of the smart positioner, it was preferable for the sake of visibili

23、ty and maintainability that any reversal be done in the control room rather than at the valve. It is important to standardize BEST OF THE BASICS 5on the location of the signal reversal to ensure that it is done and done only once. Table 1 sum-marizes how the controller action depends upon both the p

24、rocess and valve actions and on the signal reversal.Which brings us to rule of thumb number one.Table 1 Controller ActionProcess ActionVa l v e ActionSignal ReversalController ActionDirect Increase-OpenNo ReverseReverse Increase-OpenNo DirectDirect Increase-CloseYes ReverseReverse Increase-CloseYes

25、DirectDirect Increase-CloseNo DirectReverse Increase-CloseNo Reverse6 BEST OF THE BASICSRule 1 The controller action should be the opposite of the pro-cess action unless there is an increase-to-close (fail-open) con-trol valve for which there is no reversal of the valve signal. This means that you s

26、hould use reverse and direct-acting controllers for direct and reverse-acting processes, respectively. The valve signal can be reversed for a fail-open valve at many places, but it is best done in the AO block of the control system. 1.3 Controller la ModeThe names for the operational modes of the PI

27、D vary from manufacturer to manufacturer. Thus, the FOUNDATION Fieldbus modes listed next provide a uniformity that can be appreciated by all.Auto (automatic) The operator locally sets the set point. PID action is active (closed loop). In older systems, this mode is also known as the local mode. Cas

28、 (cascade) The set point comes from another loop. PID action is active (closed loop). This is also known in older systems BEST OF THE BASICS 7as the remote mode or remote set point (RSP).LO (local override) PID action is sus-pended. The controller output tracks an external signal to position the val

29、ve. This mode is typically used for auto tuning or to coordinate the loop with interlocks. In older systems, it is also known as output tracking.Man (manual) The operator manually sets the output. PID action is suspended (open loop).IMan (initialization manual) PID action is suspended because of an

30、interruption in the forward path of the controller output. This is typically caused by a downstream block that is not in the cascade mode. The controller output is back-calculated to pro-vide bumpless transfer. RCas (remote cascade) The set point is remotely set, often by another computer. PID actio

31、n is active (closed loop). This mode is also known in older systems as the supervisory mode.8 BEST OF THE BASICSROut (remote output) The output is remotely set, often by a sequence or by another computer. PID action is suspended (open loop). In older systems, this mode is known as direct digital con

32、trol (DDC).1.4 Is That Your Final Response?The contribution that the proportional action makes to the controller output is the error multi-plied by the gain setting. The contribution made by the integral action is the integrated error mul-tiplied by the reset and gain setting. The reset setting is r

33、epeats per minute and is the inverse of integral time or reset time (minutes per repeat). FOUNDATION Fieldbus will standardize the reset time setting as seconds per repeat and the rate (derivative) time setting as seconds. The contribution made by derivative mode is the rate of change of the error o

34、r process variable in percent (%), depending upon the type of algo-rithm, multiplied by the rate and gain settings. When derivative action is on the process variable instead of on the control error, it works against a set point change. (The control error is the difference between the process variabl

35、e and the set point.) The reason for this is that rate BEST OF THE BASICS 9doesnt know the process variable should be changing initially and that the brakes should only be applied to the process when it approaches set point. Using derivation action that is based on the change in control error will p

36、rovide a faster initial takeoff. This is particularly advantageous for set points driven for batch control, advanced control, or cascade control. The improvement can translate into shorter cycle or transition times, an enhancement of the ability of slave loops to mitigate upsets, and less off-spec p

37、roduct because of less overshoot. The change in controller output for a set point change is in the opposite direction of the response to process variable change. When the process variable starts to respond to the set point change, the contribution from proportional and derivative modes changes direc

38、tion. The contribution of the integral mode does not change direction until the process variable crosses the new set point leading to overshoot. When derivative action is used with a time con-stant there is a built-in filter that is about one-eighth (1/8) of the rate setting. However, you should use

39、 set point rate of change limits to pre-vent a jolt to the output when there is a large 10 BEST OF THE BASICSstep change in error from a manually entered set point. This is particularly important when you are using large gains or derivative action based on control error.The PID algorithm uses percen

40、tage (%) input and output signals rather than engineering units. Thus, if you double the scale span of the input (error or process variable), you effectively halve the PID action. Correspondingly, if you double the scale span of the output (manipulated vari-able), you double the PID action. In field

41、bus blocks, both input and output signals can be scaled with engineering units. Using output sig-nal scaling will facilitate the manipulation of the slave loops set point for cascade control. For controllers that use proportional band, you need to divide the proportional band into 100 percent to get

42、 the equivalent controller gain. Propor-tional band is the percentage change in error needed to cause a 100 percent change in output.The difference equations for a reverse acting ISA Standard Form PID with set point weight factors for the proportional mode () and derivative mode () set by PID struct

43、ure are as follows where the filter time on the derivative mode is a BEST OF THE BASICS 11fraction alpha () of the rate time (alpha is PID parameter whose default is typically 0.125):%COn= Pn+ In+ Dn+ %COiPn= Kc* ( * %SPn %PVn)The difference equations for a reverse acting Series Form PID with the sa

44、me structure options:%COn= P + I + %COiP = * ( * %SPn %PVn )InKcTi- * %SPn%PVn() * txIn 1+=DnKc*Td* *%SPn%SPn 1()%PVn%PVn 1()*Td*Dn 1+*Td*tx-=DnDnDnKcDnIDnKcTi- * %SPn%PVn Dn() * txIDn 1+=12 BEST OF THE BASICSIf the structure used does not have integral action, the term %COiis an adjustable bias and

45、 integral term is zero are not used.where%COi= controller output at the transition from MAN or ROut modes (%)%COn= controller output for current scan n (%)%COn1= controller output for last scan n1 (%)%PVn= process variable for current scan n (%)%PVn1= process variable for last scan n1 (%)%SPn= set p

46、oint for current scan n (%)%SPn1= set point for last scan n1 (%)Kc= controller gain for ISA Standard Form (%/%) (dimensionless)= controller gain for Series Form (%/%) (dimensionless)Ti= integral time (reset time) for ISA Standard Form (seconds)= integral time (reset time) for Series Form (seconds)Dn

47、Kc*Td* *%SPn%SPn 1()%PVn%PVn 1()*Td*Dn 1+*Td*tx-=KcTiBEST OF THE BASICS 13Td= derivative time (rate time) for ISA Standard Form (reset time) (seconds)= derivative time (rate time) for Series Form (seconds)Pn= proportional mode contribution for ISA Standard Form for current scan n (%)Pn1= proportiona

48、l mode contribution for ISA Standard Form for last scan n1 (%)P=proportional-derivative mode contribution for Series Form for current scan n (%)P=proportional-derivative mode contribution for Series Form for last scan n1 (%)In= proportional mode contribution for ISA Standard Form for current scan n (%)In1= proportional mode contribution for ISA Standard Form for last scan n1 (%)I=integral-derivative mode contribution for Series Form for current scan n (%)I=integral-derivative mode contribution for Series