A Short Course on Transport network analysis

ur2young 在 10/03/2021 - 10:32 提交

Name of the Lecturer: Prof. Benjamin Heydecker, University College London

Short Bio: After his graduation on the mathematical tripos at Cambridge University, Benjamin Heydecker joined the Centre for Transport Studies at UCL (University College London) where he studied for his doctorate in transport studies under the supervision of Professor Richard Allsop. He moved to the Institute for Transport Studies at the University of Leeds as Research Officer where he worked for three years. He then returned to the Centre for UCL as Lecturer in Transport Studies, where he is now Emeritus Professor of Transport Studies.

Heydecker specialises in the development and application of mathematical and statistical analysis in transport studies. His current research is in application areas of vehicle emissions as they affect urban air quality, railway traffic control, transport safety, discrete choice modelling and vehicle automation. He has particular interests in network modelling, traffic management and control, and dynamic modelling. The emerging methodology of information and communications technology (ICT) provides new opportunities for the applications that he explores.

Subject Description:

The course in Transport Network Analysis will explore the role of this quantitative approach in the transport planning process. The framework for this is the 4-stage transport planning model comprising Trip generation, Trip distribution, Mode split and Traffic assignment. Within this framework, the focus of network analysis is in the traffic assignment stage, which takes an origin-destination trip matrix as a representation of travel demand and a transport network as a representation of supply. The relationship between supply and demand is resolved by adopting an assignment principle that represents the route choice of individuals. This course will explore the relationships among supply, demand and assignment principle, and will show how the resulting model for the 4th stage of the transport planning process integrates with other parts of it. Attention will be paid to each of equilibrium, stochastic equilibrium and system optimal assignment principles. Combined transport models will also be explored that represent the integration of the assignment process together travel demand as represented by mode choice and by trip distribution to achieve mutual consistency between these elements of the transport planning process. Integration of the assignment process with network design will be introduced as a way of anticipating travellers’ response to changes in transport provision: this will be illustrated with examples to show the importance for strategic investment in transport facilities. The topic of dynamic traffic assignment will also be introduced, including the choice of travellers’ departure times.

This short course will prepare students to engage more fully in transport planning, with a clear understanding of the methods used to model transport systems, so to analyse their operation more effectively and ultimately to design them better. The broad coverage will also prepare students for future developments in this important area of transport studies.

Learning Outcomes:

After completing this course the students will be able to:

-Understand the four-stage approach to transport planning

-Formulate network models of travel

-Distinguish among different traffic assignment principles

-Compare network performance under different route choice principles

-Identify the need for combined transport models

-Understand the importance of statistical modelling approaches

-Be able to specify, estimate and interpret appropriate statistical models

-Represent transport and traffic management policies within the four-stage framework

-Interpret differences between model runs with different specifications

-Choose among different assignment principles according to the policy to be evaluated

-Incorporate value of time as a motivation for departure time choice

-Adopt dynamic transport models where appropriate

Curriculum:

Equilibrium network models

The purpose of these lectures is to familiarise students with the structure of the four-stage transport planning model and the way in which it represents travellers’ dimensions of choice. The focus is on the role of road traffic assignment and equilibrium modelling within this

Optimisation

The purpose of these lectures is to introduce to students the concepts of using unconstrained and constrained optimisation as equivalent formulations to equilibrium transport models, and interpret their results in terms of travel behaviour.

Combined transport model

The purpose of these lectures is to introduce to students ways in which stages of the transport planning process can be combined to achieve mutual consistency between travel behaviour and the associated costs.

Trip matrix estimation

The purpose of this lecture is to familiarise students with a range of approaches to using low-cost abundant traffic data to estimate travel demand trip matrices for use in road traffic assignment.

Dynamic traffic assignment

The purpose of this lecture is to extend static transport model to dynamic ones that can represent peak-period congestion and travellers’ response to this explicitly. This treatment includes choice of departure time.

Programme of Teaching, March 2021:

The course will be delivered in 8 blocks of 4 hours duration, including breaks and illustrative case studies. Students will be engaged in formative exercises to explore the methods presented, including numerical examples and dynamic simulation using provided software.

Day	Session 11	Session 12		Session 13		Hours
Monday 15 March	Introduction	4 stage transport planning model				3
Wednesday 17 March	Network modelling and analysis			Traffic management		3
Thursday 18 March	Optimisation methods		Example applications		Constrained optimisation	3
Friday 19 March	Equilibrium analysis			Equilibrium and optimisation		3
Monday 22 March	Regression modelling	Statistical models in transport		Network design		3
Wednesday 24 March	Demand estimation			Matrix estimation by maximum entropy (ME2)		3
Thursday 25 March	Network management	Congestion charging		Low emissions zones		3
Friday 26 March	Combined transport models: Integration within the 4-stage model				Dynamic network modelling	3
Scheduled in response to students’ need	Additional tutorial sessions for discussion, exercise answers and revision					8
Total						32