【2019年5月20日】【管理科学与工程系学术讲座】Designing Traffic Assignment Algorithms by Specific Topological Structures
发布时间:05-16-19

管理科学与工程系学术讲座

Lecture: Designing Traffic Assignment Algorithms by Specific Topological Structures

Speaker: Dr Jun Xie,Southwest Jiaotong University

Time:14:00-15:00, 20th May, 2019

Venue: Room 206, Tongji Building A

Abstract:

The user equilibrium (UE) static traffic assignment problem (TAP) has long been used as a standard tool to predict network flows. Under mild assumptions, the UETAP can be formulated as a convex optimization program. Designing efficient solution algorithms for this problem in large regional-scale networks has been a recurring research theme in transportation science and has attracted much attention in the past decades. The recent literature observes that the development of advanced algorithms for the TAP heavily relies on the proper use of some specific topological structures. This presentation will perform an extensive analytical and numerical investigation of most known TAP algorithms by examining how they create or utilize the specific topological structures in their algorithmic design and how these strategies will affect the convergence efficiency of the algorithms. The insights obtained from these discussions might be helpful for the potential improvement of existing algorithms or the proposition of new algorithms for large-scale network equilibrium problems.

Bio:

Jun Xie is currently an Associate Professor in the School of Transportation and Logistics, Southwest Jiaotong University. He received his Bachelor degree from Jilin University and PhD from Tongji University. He did post-doc research in Shanghai Jiaotong University and Northwestern University from 2013 to 2017. His research focuses on modellings and algorithms of traffic and transit assignment problems.Most of his papers were published in leading international journals of transportation area, such as Transportation Research Part A/B/C and Transportation Science. He received the Stella Dafermos Best Paper Award by the Network Modelling Committee of TRB in 2018.