TABASCO Del. 8.1

Telematics Applications in BAvaria, SCotland and Others

Project number TR1054

Deliverable number 08.1

Deliverable type: Final

Deliverable due date: August1997

Date: August 1997

Workpackage number: 08

Workpackage title: Urban Integrated Traffic Control

Nature of deliverable: Final

Telematics Applications in BAvaria, SCotland and Others

Project number TR1054

Deliverable number 08.1

Deliverable type: Final

Deliverable due date: August 1997

Date: August 1997

Sites, User Needs and Traffic Model Verification

Report

Workpackage number: 08

Workpackage title: Urban Integrated Traffic Control

Nature of deliverable: Final

Tom McLean, Glasgow City Council

Colin Brader, Atkins Wootton Jeffreys

Christina Diakaki, Technical University of Crete

Markos Papageorgiou, Technical University of Crete

Stefan Hangleiter, Steierwald Schφnharting und Partner

Maria Tsavachidis, Technische Universitδt Mόnchen

Christophe Damas, CΙTΙ de Lyon

Brian Maxwell, DoE (Northern Ireland)

ABSTRACT

TABASCO WP8 “Urban Integrated Traffic Control” aims to develop, validate, simulate and demonstrate integrated traffic control, harmonising motorway and UTC control techniques across the interface between inter-urban and urban environments. WP8 addresses the technical performance of integrated control techniques, their network impacts and socio-economic effects, legal/institutional questions and user acceptability.

This report focuses on the requirements, nature and verification of traffic models designed to support the development of integrated urban traffic control strategies.

The work involves demonstrations in Munich, Glasgow and Lyon. Belfast and Edinburgh are active follower sites. All of the sites comprise a mix of motorway and urban arterial streets which suffer from both recurrent and incident-related congestion.

Control models, and the strategies to be derived, are based upon a definition of user needs with respect to a list of user groups common across all sites. This report compares the approach to traffic modelling between the sites in terms of approach and models used.

The validation of the methods employed and the results obtained in Munich, Glasgow and Lyon are described. It is concluded that the models derived are sufficiently robust to be used in the definition of control strategies.

Keyword list: User Needs, Model, Validation, Munich, Glasgow, Lyon

EXECUTIVE SUMMARY

General

This report focuses on the requirements, nature and validation of traffic models designed to support the development of integrated urban traffic control strategies.

At the sites, there is a considerable history of attention to traffic problems, policies and systems at the urban/inter-urban interface. Whilst the details and history associated with each site are unique, there are a number of parallels between the sites (in terms of traffic characteristics, jurisdiction, control systems, policy and research and development) which provide some confidence that the work in WP8 can be expected to have general validity.

User Needs

To a substantial degree the user needs analysis has drawn on the accumulated wisdom of the relevant network authorities and on conventional practice in relation to traffic control systems.

Four user groups have been identified:

	network implementation authorities;
	road users;
	non road-users; and
	traffic control centre operators.

User needs derived from these groups indicate that control strategies:

	need to be sensitive to a range of efficiency, environmental and safety considerations;
	need to be considered in advance by the various authorities involved; and
	need to be operated in a way which obtains reliable control with minimal effort and satisfactory response times.

Traffic models need to enable:

	potential impacts to be assessed as part of the strategy development process;
	strategies to be demonstrated to and agreed with relevant authorities; and
	the strategy implementation process to be as straightforward and direct as possible.

Traffic Models

The approach to traffic modelling differs between the sites, in a number of respects.

In Munich, traffic models are built into a set of interrelated control tools which are linked through a defined system architecture using established data exchange procedures. Development of control strategies involves the running of the control process in simulation mode and adjusting relevant parameters until a satisfactory outcome is achieved. In Glasgow and Lyon, the function of the traffic model is different, their use in both cases being restricted to the off-line development and testing of control strategies. Thus in Glasgow and Lyon, traffic models are not run as part of the control strategy implementation process.

The nature of the models also shows variation:

	both Munich and Scotland adopt a macroscopic simulation of the network (variants of METANET) to analyse the possible implications of re-routeing;
	Lyon, on the other hand, uses a micro simulation model for this purpose;
	a second network-wide model is used in Munich for the assessment of the full range of socio-economic effects of possible control strategies. This second model also assists with the overall evaluation of control strategies in Munich.

Model Verification

Verification reflects the different modelling approaches adopted at each site. Within both Munich and Glasgow quantitative and qualitative methods were used. This has involved a comparison of modelled absolute flow and delay values against observed or detected values, and, a comparison of queue and diversion effects. Within Lyon the validation of the INTEGRATION model has concentrated upon quantitative validation of vehicle flow with the object of accuracy within 5%.

Within Munich the AIDA, incident detection, and NEMO assignment and evaluation, models have been validated separately. AIDA has been validated during a field trial of off line operation. This showed the accurate detection of all incidents according to the system requirements. The NEMO model was validated according to junction performance, routeing and absolute flow levels. It was shown that the model functioned with sufficient accuracy.

The Glasgow, METACOR, model was subjected to quantitative validation of the motorway network and a qualitative validation of the urban network. It was concluded that the results were acceptable in terms of achieved accuracy of the model calculations, the representation of traffic conditions on different days, and representation of observed congestion conditions.

Conclusion

At each demonstration site it was concluded that the models constructed were sufficiently detailed and capable of realistically describing traffic phenomena at the respective sites thus allowing the definition and testing of control strategies for implementation. The production of control strategies will be described in TABASCO deliverable 8.2.

Table of Contents

1. INTRODUCTION

1.1 The TABASCO Context

1.2 Control Models in Context

1.3 This Report

1.4 The Test Sites

2. THE SITES - PROBLEMS, POLICIES AND INFRASTRUCTURE

2.1 Summary

2.2 Munich

Site and Problem Description

Transport Policy: Context

Transport Policy: Implementation

2.3 Glasgow

Site and Problem Description

Transport Policy: Context

Transport Policy: Implementation

Infrastructure

2.4 Lyon

Site and Problem Description

Transport Policy: Context

Traffic Policy: Implementation

Infrastructure

2.5 Belfast

Site and Problem Description

Transport Policy: Context

Infrastructure

Role Within TABASCO

3. USER NEEDS

3.1 General Approach

3.2 User Groups

Implementation Authorities

Road Users

Non Road-Users

Traffic Control Centre Operators

3.3 User Needs

Implementation Authorities

Road Users

Non Road-Users

Traffic Control Centre Operators

3.4 Functional Requirements of Control Strategies and Models

4. MODEL DESCRIPTIONS

4.1 Overview

4.2 Munich

Introduction

TIP / SSC

AIDA

VARIA

STRAMO

NEMO

STRAMA

Interplay of NEMO and STRAMA

4.3 Glasgow

METACOR Modelling Approach

Network Representation

Definition of Traffic Variables

METACOR Model Equations

Control Consideration within METACOR

METACOR Application to the Glasgow Site

4.4 Lyon

Principles of the Model

Functions of the Model

Model functional architecture

5. MODEL VERIFICATION AND SUITABILITY

5.1 Introduction

5.2 Munich

AIDA

NEMO

The Next Steps

5.3 Glasgow

Introduction

Quantitative Model Validation

Model Application to the M8 Eastbound Motorway

Parameter Estimation and Model Validation

Conclusion

Qualitative Model Validation

The Qualitative Validation Procedure

Conclusion

The Next Steps

5.4 Lyon

Introduction

First level of validation

Second level of validation

Summary

5.5 Model Validation Conclusion

6. CONCLUSION

REFERENCES

Appendix A Glasgow METACOR validation procedure

Basic Modelling Equations

The Identification Procedure

Appendix B Glasgow METACOR: Validation mean speed/flow trajectories & space time maps