Simulation based hybrid model for a partially automatic dispatching of railway operation [Elektronische Ressource] = (Simulationsbasiertes Hybridmodell für eine teilautomatisierte Disposition des Eisenbahnbetriebs) / vorgelegt von Yong Cui
172 Pages
English
Downloading requires you to have access to the YouScribe library
Learn all about the services we offer

Simulation based hybrid model for a partially automatic dispatching of railway operation [Elektronische Ressource] = (Simulationsbasiertes Hybridmodell für eine teilautomatisierte Disposition des Eisenbahnbetriebs) / vorgelegt von Yong Cui

-

Downloading requires you to have access to the YouScribe library
Learn all about the services we offer
172 Pages
English

Informations

Published by
Published 01 January 2010
Reads 24
Language English
Document size 3 MB

Exrait

Simulation-Based Hybrid Model for a Partially-
Automatic Dispatching of Railway Operation
(Simulationsbasiertes Hybrid Modell für eine
teilautomatisierte Disposition des Eisenbahnbetriebs)

Von der Fakultät Bau- und Umweltingenieurwissenschaften
der Universität Stuttgart zur Erlangung der Würde eines
Doktors der Ingenieurwissenschaften (Dr.-Ing.) genehmigte Abhandlung


vorgelegt von

Yong Cui
aus Shitai, China

Hauptberichter: Prof. Dr.-Ing. Ullrich Martin
Mitberichter: Prof. Dr.-Ing. Jörn Pachl
Tag der mündlichen Prüfung: 07.12.2009



Institut für Eisenbahn- und Verkehrswesen der Universität Stuttgart
2010



Foreword
The use of software-based simulation models for infrastructure dimensioning and
scheduling has become common practice within the last few years due to the enor-
mous increase in the performance capabilities of both computer hardware and soft-
ware.
However, the practical application of these models had been restricted due to the
lack of:
• theoretical approaches regarding the advantageous, integrated interaction of
asynchronous and synchronous simulation
• consideration to the specific requirements of shunting movements
• real-world-applicable solutions for avoiding deadlocks
One of the most important results of the research in this thesis is the design of a
comprehensive, hybrid simulation model based on the current knowledge of the syn-
chronous and asynchronous simulation of railway operational processes that can be
used for the dimensioning of infrastructures, scheduling, as well as during the auto-
mated dispatching of railway operation.
Until now, there had been no solutions to the problem of the gaps existing between
the macroscopic and microscopic models.
With this model, a standardised application for the simulation of railway operations in
applications ranging from extensive networks to shunting services is now possible.
This is an important prerequisite for the implementation of automatic train and shunt-
ing dispatching solutions.
The improved capability of the modelling and the further automation of the extensive
network simulation at different levels are particularly commendable innovations
achieved in this thesis.
Furthermore, the consideration given to specific issues, the detailed discussion of
topic-related facts, the objective evaluation of individual influence factors, as well as
the outstanding methodical combination of the components of the model itself are
equally commendable.

This thesis represents a significant advance in the state of knowledge in the scientific
field of transportation modelling and can furthermore be used in other comparable
fields where the subject of deadlock avoidance requires special attention.





Dr.-Ing. Ullrich Martin
Universitätsprofessor


Vorwort
Die Anwendung von softwaregestützten Simulationsmodellen zur Dimensionierung
von Infrastrukturen und zur Fahrplanerstellung wurde in den letzten Jahren durch die
erheblich verbesserten rechentechnischen Möglichkeiten zum Standard. Allerdings
schränken die hier eingesetzten Modelle aufgrund fehlender
• modelltheoretischer Ansätze zum vorteilhaften Zusammenspiel asynchroner
und synchroner Simulationslösungen für deren integratives Zusammenwirken,
• Integration der besonderen Bedingungen des Rangierdienstes sowie
• praxisorientierter Lösungen zum Vermeiden von Deadlocks
den praktischen Einsatz ein oder erschweren diesen signifikant.
Wesentliches Forschungsergebnis der vorliegenden Arbeit ist der Entwurf eines um-
fassenden hybriden Simulationsmodells auf der Grundlage der aktuellen Erkennt-
nisse der synchronen und asynchronen Eisenbahnbetriebssimulation, das sowohl
betriebsvorbereitend im Rahmen der Infrastrukturdimensionierung und Fahrplaner-
stellung als auch im Eisenbahnbetrieb bei der teilautomatisierten Disposition zum
Einsatz kommen kann. Bislang noch vorhandene Lücken im Zusammenspiel zwi-
schen makroskopischer und mikroskopischer Betrachtung und die praxisorientierte
Berücksichtigung von Deadlocks werden durch neuartige Lösungen geschlossen.
Damit wird mit dem vorgeschlagenen Modell auch eine einheitliche Anwendung der
Eisenbahnbetriebssimulation von der großräumigen Netzbetrachtung bis hin zum
Rangierdienst möglich und somit eine wichtige Voraussetzung für eine selbsttätige
Zug- und Rangierlenkung geschaffen.
Die Verbesserung der Modellierung und die weitere Automatisierung bei Netzbe-
trachtungen sind besonders anerkennenswerte Neuerungen. Die Berücksichtigung
der einzelnen Fragestellungen, die ausführliche Auseinandersetzung mit verschiede-
nen, die Thematik tangierenden Sachverhalten und eine objektive Wertung einzelner
Einflußfaktoren sowie die hervorragende methodische Zusammenführung der einzel-
nen Komponenten des Modellkomplexes sind besonders hervorzuheben. Die vorlie-
gende Dissertation stellt eine signifikante Erweiterung des Wissensstandes der ver-
kehrswissenschaftlichen Modellierung im Bereich der Simulation dar und kann darü-

ber hinaus auch in anderen vergleichbaren Gebieten, in denen die
Deadlockvermeidung besonderer Aufmerksamkeit bedarf, zum Einsatz kommen.





Dr.-Ing. Ullrich Martin
Universitätsprofessor





Acknowledgement
It is a great honor for me to thank those who supported and inspired me throughout
my dissertation work in the past years.
I would like to gratefully acknowledge the help given by Prof. Dr.-Ing. Ullrich Martin.
Motivated from his research, I was encouraged to explore the solution in the field of
automatic dispatching and deadlock avoidance. During the research, he gave me
valuable and constructive suggestion and advice that enrich the algorithm and my
idea. Meanwhile he also demonstrated me how to carry out the research work in a
structural and effective way.
Prof. Dr.-Ing. Jörn Pachl has provided me with a comprehensive understanding on
the theory and the state of art of deadlock avoidance. I appreciate him agreeing to be
my co-supervisor and his insightful comments on my work.
It is a great opportunity to express my thanks to my colleagues, who gave me plenty
of support and help in the past projects. Dr.-Ing. Harry Dobeschinsky and Stefan
Trischler helped me to get a good understanding on the workflow of dispatching and
the traffic simulation in project RUDY. Macro Neuber introduced me the development
of railway dispatching and gave me a lot of support in my master thesis and in project
PULZURE. Working with Bernd Raubal in project PULRAN, I gained the experience
of synchronous simulation and deadlock avoidance. These projects and experiences
are the foundation of my dissertation work.
During the writing of the dissertation, Maureen Lynch gave me a lot of help and sug-
gestion on the way for good expression and my English. Teresa Krohn helped me to
improve and standardize the formatting of mathematic formula and literature refer-
ence. I am also thankful for their carefully reading and commenting on my work.
Finally I would like to thank my wife Jing Fu. It is impossible for me to concentrate
and enjoy my work without her unconditional support, encouragement and under-
standing.







Table of Contents



Table of Contents
Index of Figures 12 
Index of Tables 14 
Abstract 15 
Zusammenfassung 16 
1 Introduction 17 
2  Railway Operation Control and Automatic Dispatching 20 
2.1  Dispatching in Railway operation Control System 20 
2.1.1  Railway Traffic Control 20 
2.1.2  Dispatching Process in Railway operation Control 23 
2.1.3  Measures of Railway Dispatching 24 
2.1.4  PULZURE: a Software Tool for Guiding Train Movements 26 
2.2  Research and Applications on the Field of Automatic Dispatching 28 
2.2.1  Simulative Models 29 
2.2.2  Analytical Models 31 
2.2.3  Heuristic Models 33 
2.3  Simulation Based Hybrid Approach for Automatic Disposition 35 
3  Synchronous Simulation in Railway Operation 37 
3.1  The Components of Synchronous Simulation 37 
3.1.1  Infrastructure Resources 38 
3.1.2  Simulation Performers 40 
3.1.3  Simulation Tasks 41 
3.2  The Workflow of Synchronous Simulation 44 
3.2.1  Initialization and Termination of Simulation 44 
3.2.2  A Single Processing Step of Synchronous Simulation 45 

Simulation-Based Hybrid Model for a Partially-Automatic Dispatching of Railway Operation 9
Table of Contents


3.3  Event-Driven Simulation 51 
3.3.1  The Principle of Event-Driven Simulation 51 
3.3.2  Event List and Event Processing in Synchronous Simulation 52 
4  Resolving Deadlocks in Synchronous Simulation 55 
4.1  Deadlock Problem and Approaches for Resolving Deadlocks 56 
4.2  Algorithms for Deadlock Avoidance in Railway Operation 59 
4.2.1  Movement Consequence Analysis (MCA) 59 
4.2.2  Dynamic Route Reservation (DRR) 62 
4.2.3  Petersen and Taylor Algorithm 64 
4.2.4  Labeling Algorithm and Single-path Problem 65 
4.3  The Banker’s Algorithm 65 
4.3.1  The Principle of the Banker’s Algorithm 66 
4.3.2  Banker’s Algorithm in Railway Synchronous Simulation 68 
4.3.3  Examples of the Banker’s Algorithm 72 
4.4  Improvements to the Banker’s Algorithm and System Performance 79 
4.4.1  Analysis with Potential State Transitions 81 
4.4.2  Test Processes in a Right Order 86 
4.4.3  Alternative Route 88 
4.4.4  Timing of State Tests 93 
4.5  Software Implementation and Evaluation for Synchronous Simulation
and Deadlock Avoidance 95 
5  Train Priorities and Allocating Infrastructure Resources in
Simulative Dispatching 98 
5.1  Train Priorities in Dispatching 100 
5.1.1  General Principles for Determining Train Priorities 100 
5.1.2  Calculation of Priority Values 101 
5.1.3  An Example for Calculating Train Priorities 104 
5.2  Static Resource Allocation and Dynamic Resource Requesting 107 

10 Simulation-Based Hybrid Model for a Partially-Automatic Dispatching of Railway Operation