ISA MDLNG CNTRL SIM DIAG CMPLX IND-2009 Modeling Control Simulation and Diagnosis of Complex Industrial and Energy Systems.pdf

1、MODELING, CONTROL, SIMULATION, AND DIAGNOSIS OF COMPLEX INDUSTRIAL AND ENERGY SYSTEMSANIPLA.book Page i Thursday, January 15, 2009 10:26 AMANIPLA.book Page ii Thursday, January 15, 2009 10:26 AMMODELING, CONTROL, SIMULATION, AND DIAGNOSIS OF COMPLEX INDUSTRIAL AND ENERGY SYSTEMSEdited byLuca Ferrari

2、ni and Carlo VeberANIPLA.book Page iii Thursday, January 15, 2009 10:26 AMModeling, Control, Simulation, and Diagnosis of Complex Industrial and Energy SystemsCopyright 2009 by ISAInternational Society of Automation67 Alexander DriveP.O. Box 12277Research Triangle Park, NC 27709All rights reserved.P

3、rinted in the United States of America.1098765432ISBN-13: 978-1-934394-90-8No part of this work may be reproduced, stored in a retrieval system, or transmitted in any formor by any means, electronic, mechanical, photocopying, recording or otherwise, without theprior written permission of the publish

4、er.NoticeThe 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 bythe reader, both the author and the publisher disclaim any and all liability of any kind arisingout of such

5、 use. The reader is expected to exercise sound professional judgment in using any ofthe information presented in a particular application. Additionally, neither the author nor thepublisher have investigated or considered the effect of any patents on the ability of the reader touse any of the informa

6、tion in a particular application. The reader is responsible for reviewingany possible patents that may affect any particular use of the information presented.Any references to commercial products in the work are cited as examples only. Neither theauthor nor the publisher endorses any referenced comm

7、ercial product. Any trademarks ortradenames referenced belong to the respective owner of the mark or name. Neither the authornor the publisher makes any representation regarding the availability of any referencedcommercial product at any time. The manufacturers instructions on use of any commercialp

8、roduct must be followed at all times, even if in conflict with the information in this publication.Library of Congress Cataloging-in-Publication DataModeling, control, simulation, and diagnosis of complex industrial and energy systems / edited by Luca Ferrarini and Carlo Veber.p. cm.ISBN 978-1-93439

9、4-90-81. Plant engineeringSimulation methods. 2. Electric power systemsAutomatic control. 3. MachinerySimulation methods. 4. Manufacturing processesAutomation. I. Ferrarini, Luca. II. Veber, Carlo.TS184.M628 2009670.42dc222008051680 ANIPLA.book Page iv Thursday, January 15, 2009 10:26 AMTo Gabriella

10、, who has provided excellent administrative support toANIPLA over many years. She has done so with such a personal touch,such compassion and care that stems from knowing that deadlines arefar easier to meet when personal relationships are paramount.ANIPLA.book Page v Thursday, January 15, 2009 10:26

11、 AMANIPLA.book Page vi Thursday, January 15, 2009 10:26 AMviiThe O3neida Publications Series This book is one of a series of books to be produced within O3neida on varioussubjects related to distributed automation.O3neida is a Canadian not-for-profit corporation. O3neida Europe is a not-for-profit a

12、ssociation headquartered in Brussels, Belgium. Together they formthe hub of the O3neida networks. Their joint mission is to operate as a networkof networks fostering the development and deployment of distributed industrialautomation technologies based on open standards. These standards include,among

13、 others, the Foundation for Intelligent Physical Agents (FIPA), theDevice Profile for Web Services (DPWS), Web Crawler (WC), and Interna-tional Electrotechnical Commission (IEC) 61131 and 61499.This book contains a collection of papers focused on the simulation, diag-nosis, modeling and predictive c

14、ontrol of complex industrial systems. Thesepapers have been developed under the aegis of ANIPLA, the national automa-tion association of Italy, which celebrated its 50th anniversary in 2006.ANIPLA, Associazione Nazionale Italiana Per LAutomazione, thenational automation association of Italy, was fou

15、nded in 1956. It is a scientificnon-profit association, whose mission is to promote the discussion, dissemina-tion and education on the technological, industrial, application-oriented andsocial aspects related to automation. Its main technical services include organi-zation of workshops, courses, co

16、ngresses, journal papers, and books.ANIPLA is the meeting point of different competencies and know-howstemming from Industry, Research Institutes and Universities across a widerange of different application areas. ANIPLAs overall objective is to give com-petitive advantage to the extent possible, to

17、 automation-related enterprises inall Italian industrial sectors. Within this book, there is a strong focus on the real-time monitoring andcontrol of energy production and delivery systems and on the application ofinnovative approaches ranging from the predictive control of a gasoline engine,through

18、 fuzzy inference applied to quality control in the paper industry and upto innovative load shedding and demand management in national electricalgrids. This book will be of interest to practitioners within the automation field,particularly those focused on energy systems. It will also be of interest

19、to aca-demics and students seeking an overview of current thinking in this field orlooking for detailed treatment of any of the issues covered by the individualchapters.ch00a2_O3 desc.fm Page vii Thursday, January 15, 2009 10:41 AMviii The O3neida Publications Series More than forty authors from cou

20、ntries around the world have contributedto the production of this unique book and O3neida thanks them, one and all, fortheir strong collaboration in producing this excellent compendium and for theircontinuing contribution to the advancement of automation process control.Future volumes in the O3neida

21、/ISA series on automation will addressother equally pressing issues such as Ethernet Safety Devices and also Ontolo-gies. O3neida will also publish materials on automation objects as part of thisseries.Finally, this book is the result of a concerted effort by many O3neidamembers. I thank them all fo

22、r their dedication and commitment to O3neida asvolunteers. I particularly thank Luca Ferrarini from Polimi in Milan, Italy forleading this effort and Allan Martel, O3neida Chief Operating Officer, for coor-dinating and managing the development of the O3neida series of books on dis-tributed automatio

23、n.I also thank ISA for their interest and support in making the publicationand distribution of this important book possible.Antonio ValentiniChief Executive OfficerO3neidach00a2_O3 desc.fm Page viii Thursday, January 15, 2009 10:41 AMixTable of ContentsList of Figures xviiList of Tables xxv1 Remote

24、Supervision Center for Enel Combined Cycle Plants 1Introduction 1Location of the center 2Architecture 2Functions 6Performance control 6Heat rate evaluation 6Maximum power forecast 7Plant status and status monitor 10Plant start-up: technical and economical evaluation 11Power unbalance calculation 14D

25、iagnostics 15Automatic reporting 20Heat rate losses 20Start-up evaluation 22Energy unbalance 25Gas turbine output temperatures and humming and acceleration phenomena 27Gas turbine compressor filters status 28Computerized events register 31Acknowledgments 32References 322 Pickling Line Modeling for A

26、dvanced Process Monitoring and Automation35Introduction 35Pickling of carbon steel 35Pickling of stainless steel 36Management and control of pickling processes 38Advances in pickling line automation 39Architecture of control software 39Pickling lines components and configuration 41ANIPLA.book Page i

27、x Thursday, January 15, 2009 10:42 AMx Table of ContentsMain components of pickling lines 42Pickling lines configuration 43Electrolytic pickling lines 43Pickling line model 44Equations describing the recirculation tank 44Equations describing the working tank 46The pickling model 48Electrolytic pickl

28、ing model 48Additional notes on the pickling line model 49Model implementation 49Conclusion 49Acknowledgments 51References 523 Modeling, Simulation and Predictive Control of a Gasoline Engine53Introduction 53Mean value engine model 55Air supply system 55Engine 58Vehicle model 60Validation 61Control

29、design 61Design of a static regulator 64Model of the driver 65Design of a dynamic controller with MPC 65Simulation results 67Conclusion 68Acknowledgments 72References 724 Dynamic Principal Component Analysis Applied to the Monitoring of a Diesel Hydrotreating Unit 75Introduction 75Hydrotreating Unit

30、 Model 76Hydrotreating (HDT) unit 76HDT unit modeling 77Principal Components Analysis (PCA) 82Monitoring system: development and results 84ANIPLA.book Page x Thursday, January 15, 2009 10:42 AMTable of Contents xiOperational conditions 84DPCA: definition of the number of delays 87DPCA: training 88A

31、hybrid procedure combining DPCA and classification 90Results: validation and test 91Conclusion 93Acknowledgments 94References 945 A Simulation Study of the Flue Gas Path Control System in a Coal-Fired Power Plant 97Introduction 97The plant model 98Structure of the plant 98Unit modeling 99The control

32、 system model 104General remarks 104Control system architecture 105Continuous-time controllers 106Logic controllers 106Improvement of the control strategy 106Improvement of the critical logic control behavior 109Selected simulation results 110Load dispatching 110Transition from FGD inserted to FGD b

33、ypassed 111Conclusion 112References 1146 Automatic Diagnosis of Valve Stiction by Means of a Qualitative Shape Analysis Technique.115Introduction 115Valve stiction 116Automatic detection of stiction 118Techniques based on PV-OPbrief review 118Techniques based on qualitative description formalism 119

34、The Yamashita stiction detection technique 121Application on simulated data 123Noise-free data 125Adding noise 125ANIPLA.book Page xi Thursday, January 15, 2009 10:42 AMxii Table of ContentsVarying setpoints 126First conclusions about the technique 129Application to plant data 129Results 130Sampling

35、 time 134Observation window 134Noise level 135Other phenomena observed in the plant data 135Conclusion 135References 1367 Monitoring and Controlling Processes with Complex Dynamics Using Soft Sensors139Introduction 139Case study 1: freeze-drying of pharmaceuticals 140Detailed and simplified models 1

36、42Observers design 145Feedback temperature control 149Case study 2: catalytic combustion of lean mixtures 150Case study 3: SCR unit for NOx154Conclusions 158Acknowledgments 158Nomenclature 160References 1608 Estimation of a Ternary Distillation Column via a Tailored Data Assimilation Mechanism 163In

37、troduction 163Estimation problem 164Data assimilation mechanism 168Estimation design 175The Non-linear Geometric Estimator (NGE) 176The Extended Kalman Filter (EKF) with reduced data injection 177Conclusion 179References 180ANIPLA.book Page xii Thursday, January 15, 2009 10:42 AMTable of Contents xi

38、ii9 A Prediction Error-Based Method for the Performance Monitoring of Model Predictive Controllers 183Introduction 183Problem statement 185Process, model, and state estimator 185Steady-state target calculation 187Dynamic optimization 189Method 190Preliminary definitions of prediction error 190Motiva

39、ting example 191Prediction error-based diagnosis 191Case studies 196Extensive simulations 196An industrial example 198Conclusion 198Acknowledgments 201References 20110 An Intelligent/Smart Framework for Real-Time Process Monitoring and Supervision .205Introduction 205Integrated framework 207Trend an

40、alysis and preprocessing 207Outlier detection 207Noise reduction 209Fault detection and identification 209Self-Organizing, Self-Clustering Network (SOSCN) 215Case study 216Conclusion 22111 Quality Monitoring Through a Dynamic Neural Software Sensor.225Introduction 225Background 226Problem statement

41、227The process 227Software sensor 228Software sensor design 228Basic structure 228ANIPLA.book Page xiii Thursday, January 15, 2009 10:42 AMxiv Table of ContentsNeural software sensor formulation 230Industrial application 231Data acquisition 231Input selection 232Results and discussion 233Conclusion

42、235References 23612 Wind Generation and Flexible Electric Load Management Issues for System Operation in Crete239Introduction 239Green Electricity Availability Barometer Service (GEA BASE) 241A Control Center tool 243Demand nodes model 243Generation nodes model 244Implementation 244Formulation of th

43、e knowledge base 245Inference derivation process 247Architecture of an expert system 248Incorporation of the GEA BASE tool into GIS and digital database for the Crete Power System 250Conclusion 251References 25113 A Fuzzy Inference System Applied to Quality Control in the Paper Industry. 253Introduc

44、tion 253Problem description 255Experimental setup 255The quality control system 257The image processing phase 258Defect detection through a clustering algorithm 262Defect evaluation through a fuzzy inference system 264Numerical results 268Conclusion and future work 270References 272ANIPLA.book Page

45、xiv Thursday, January 15, 2009 10:42 AMTable of Contents xv14 Innovative Load Shedding and Demand Side Management Enhancements to Improve the Security of a NationalElectrical System .273Introduction 273Demand side management and demand response services 274Automatic meter reading system and enhancem

46、ent required by Demand Response (DR) services 277Potential vulnerability of communication technologies for demand control services 278Current activity in CESI RICERCA 279Conclusion 282Acknowledgments 283References 283Index 285ANIPLA.book Page xv Thursday, January 15, 2009 10:42 AMANIPLA.book Page xv

47、i Thursday, January 15, 2009 10:42 AMxviiList of FiguresFigure 11 The Remote Supervision Center 3Figure 12 SSCC system architecture 5Figure 13 Area display and display tree 5Figure 14 Instantaneous heat rate losses 8Figure 15 Display of average heat rate losses 9Figure 16 Plant status automata 11Fig

48、ure 17 Plant status monitor automata 12Figure 18 Display of start-up counter 14Figure 19 Start-up animated Gantt chart 15Figure 110 Area diagnostics summary display 17Figure 111 Main plant diagnostic alerts 18Figure 112 On-Off fieldbus actuator alerts 19Figure 113 Heat rate losses report 21Figure 11

49、4 Total area start-up counters and costs page 24Figure 115 Example of a plant cold start-up counters and costs page 26Figure 116 Gas turbine output temperatures report 29Figure 117 Gas turbine acceleration and humming report 30Figure 21 Different designs of pickling line tanks: (a) Deep-type tanks (b) Shallow-type tanks (c) Turbulence tanks (d) Turboflo tanks 37Figure 22 Architecture of control software 40Figure 23 Typical PWS interface 41Figure 24 Schematic representation of a pickling stage 43Figure 25 Typical pickling line configuration 44Figure 26 Pickli