Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
©2006Fisher-Rosemount Systems, Inc.Slide 1- 1
Process Control Introduction – Historic Perspective
Automation Automation Connections Connections ISA EXPO 2003ISA EXPO 2003A Practical Guide To Advanced Control
Terry Blevins
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 2
A Practical Guide To Advanced Control A Practical Guide To Advanced Control A Practical Guide To Advanced Control Presentation will provide a summary of key points and examples from Advanced Control Unleashed
Book was inspired by the latest generation of advanced control products.
The application sections include guided tours of each technology
CD provides an overview video for each section and examples.
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 3
Book ContentBook ContentBook ContentIntroductionSetting the FoundationAPC PathwaysEvaluating System PerformanceAbnormal Situation Management Automated TuningFuzzy Logic ControlProperties EstimationModel Predictive ControlVirtual Plant
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 4
Structure of each SectionStructure of each SectionStructure of each SectionPractice– Overview– Opportunity Assessment– Examples
Application– General Procedure– Application Detail– Rules of Thumb– Guided Tour
Theory
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 5
Advanced Control Unleashed Advanced Control Unleashed Advanced Control Unleashed IntroductionSetting the FoundationAPC PathwaysEvaluating System PerformanceAbnormal Situation Management Automated Tuning Fuzzy Logic ControlProperties EstimationModel Predictive ControlVirtual Plant
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 6
APC for the Average Process EngineerAPC for the Average Process EngineerAPC for the Average Process Engineer
The advent of powerful and friendly integrated software has moved advanced process control (APC) from the realm of consultants into the arena of the average process control engineer
There is a opportunity of the engineers closest to the process and daily operation to take a much more active role in the development and support of APC applications
Rule of thumb and concise explanations are employed to save the reader time to obtain specific implementation guidance.
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 7
Advanced Control Unleashed Advanced Control Unleashed Advanced Control Unleashed IntroductionSetting the FoundationAPC PathwaysEvaluating System PerformanceAbnormal Situation Management Automated Tuning Fuzzy Logic ControlProperties EstimationModel Predictive ControlVirtual Plant
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 8
A Solid Foundation is EssentialA Solid Foundation is EssentialA Solid Foundation is EssentialThe core of a solid foundation is good measurements and final elements.
Deficiencies in the measurement and final element can increase the time required for process testing and identification by a factor of 5 or more and can significantly reduce the improvement in process capacity and efficiency provided by APC.
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 9
Rules of ThumbRules of ThumbRules of ThumbThe largest opportunity for final elements if to eliminate stick-slip and deadband. The effect of slip is worse than stick and stick is worse than deadband.
The largest opportunity for measurements is the selection and installation of sensors.
Check the life-cycle cost, including the cost of variability before choosing a less expensive control valve or measurement.
Use smart transmitters. The improved accuracy and diagnostics are worth the extra cost. Use Fieldbus for major upgrades and new installations.
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 10
Advanced Control Unleashed Advanced Control Unleashed Advanced Control Unleashed IntroductionSetting the FoundationAPC PathwaysEvaluating System PerformanceAbnormal Situation Management Automated Tuning Fuzzy Logic ControlProperties EstimationModel Predictive ControlVirtual Plant
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 11
Objective – APC Pathway Objective Objective –– APC Pathway APC Pathway
Examines how to identify advanced process control opportunities, estimate benefits, select the best technology, sustain the solution, track the benefits
Industry is driven by cost and benefit analysis and is generally not interested in a great technology looking for an application.
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 12
Basic Process Control System
Loop Performance Monitoring SystemProcess Performance Monitoring System
Abnormal Situation Management SystemAuto Tuner
Fuzzy Logic
Property EstimatorsModel Predictive Control
Ramper or Pusher
LP/QPRTOTS
TS is tactical scheduler, RTO is real time optimizer, LP is linear program, QP is quadratic program
Pyramid of TechnologiesPyramid of TechnologiesPyramid of Technologies
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 13
2-Sigma 2-SigmaRCAS
Set Point
LOCALSet Point
2-Sigma2-Sigma
Upper LimitPV distribution for original control
PV distribution forimproved control
Extra margin whenoperator manually adjusts set points
value
Effect of Operator Conservatism on Proximity to ConstraintEffect of Operator Conservatism on Effect of Operator Conservatism on Proximity to ConstraintProximity to Constraint
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 14
Benefits from Operation Closer to a Controlled Variable LimitBenefits from Operation Closer to a Benefits from Operation Closer to a Controlled Variable LimitControlled Variable Limit
Before Advanced Control
$$Savings
Limit or Spec Target
Setpoint
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 15
Advanced Control Unleashed Advanced Control Unleashed Advanced Control Unleashed IntroductionSetting the FoundationAPC PathwaysEvaluating System PerformanceAbnormal Situation Management Automated Tuning Fuzzy Logic ControlProperties EstimationModel Predictive ControlVirtual Plant
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 16
Model Predictive ControlModel Predictive ControlModel Predictive ControlModel Predictive Control (MPC) has been proven to provide benefits, in the right application, that are greater than those achieved from the improvement of basic control systems.
The greatest benefits are realized in applications with dead-time dominance, interactions, constraints, and the need for some optimization.
Optimization is often as simple as the maximization or minimization of a flow.
The advantage of MPC lies in its knowledge of the effects of past actions of manipulated and disturbance variable on the future profile of the controlled and constraint variables.
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 17
SP Prediction
CV Prediction
ProcessControl
Algorithm-
-
Setpoint (SP)
Disturbance (DV)
Manipulated (MV)
Model output correction
Future CV
Future SP Error
Vector
Model output
DV
Controlled (CV)
Feedback of Future Trajectories for Model Predictive ControlFeedback of Future Trajectories for Feedback of Future Trajectories for Model Predictive ControlModel Predictive Control
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 18
Graphic illustration of the operation of a MPC controllerGraphic illustration of the operation of Graphic illustration of the operation of a MPC controllera MPC controller
time
CV
MV
Setpoint trajectory
Past Future
CV prediction
Future calculated moves
Predicted errors
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 19
FC 1-4
FT 1-4
FC 1-3
FT 1-3
FC 1-1
FT 1-1
FC 1-2
FT 1-2
LT 1-2
LC 1-2TT
1-4
TC 1-2
FC 2-1
FT 2-1
LT 1-3
LC 1-3 RSP
RSP
RSP
Distillate Receiver
Tower A
A Overheads
A Bottomsto Column B
A Steam
A Feed
AT 1-2
A Reflux A Analyzer
PC 1-4
PT 1-4
A VentAir Cooler
PT 1-2
TT 1-5
DP 1-1
TT 1-2
PT 1-3
TT 1-1
This is the measurement and control for a particular set of columns and is not intended to suggest a general scheme for most columns
Distillation Tower ExampleDistillation Tower ExampleDistillation Tower Example
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 20
Manipulated
Con
trol
led
FC 1-3Reboil Rate
AT 2-2A btms comp
TC 1-2Top Temp
AT 1-2A ovhd comp
Con
stra
int FC 1-2
Reflux rateTT 1-4Reflux tempTT 1-5Tower A temp
DisturbancePT 1-3Steam Pressure
FC 1-1Feed Rate
DP 1-1Delta P
FEED TGTA ovhd comp
Distillation Tower, Step Response Model for MPC DevelopmentDistillation Tower, Step Response Distillation Tower, Step Response Model for MPC DevelopmentModel for MPC Development
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 21
Configuration of MPC StrategyConfiguration of MPC StrategyConfiguration of MPC Strategy
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 22
Automated process testing to identify the process modelAutomated process testing to identify Automated process testing to identify the process modelthe process model
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 23
Step Response ModelStep Response ModelStep Response Model
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 24
Comparison of FIR and ARX Step Response ModelComparison of FIR and ARX Step Comparison of FIR and ARX Step Response ModelResponse Model
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 25
Verification of identified modelVerification of identified modelVerification of identified model
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 26
Testing of control using simulated environmentTesting of control using simulated Testing of control using simulated environmentenvironment
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 27
Operator interface to MPCOperator interface to MPCOperator interface to MPC
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 28
Summary Summary Summary The advent of powerful and friendly integrated software has moved advanced process control (APC) from the realm of consultants into the arena of the average process control engineer
Deficiencies in the measurement and final element can significantly reduce the improvement in process capacity and efficiency provided by APC.
Advanced process control techniques may be implemented in stages. Most plants can justify the utilization of model predictive control.
Copyright © 2003 ISA. Reprinted by permission. All rights reserved.Slide 29
QuestionsQuestionsQuestionsQuestions from the Audience?
Related Resources from ISA
– Advanced Control Unleashed, available at www.isa.org