Network Models for Connected and Autonomous Vehicles (CAV)

Learning Objectives:

1. Analyze transportation systems using game theory and optimization by drawing on analogies between the interactions of supply and demand in economic markets.
   a. Define selfish routing (user equilibrium) and system optimal problems
   b. Develop solution algorithms for large scale networks
2. Apply specialized network structures to solve selfish routing games efficiently.
3. Rigorously formulate and apply various extensions of the selfish routing game – traffic assignment with elasticity, stochastic user equilibrium, network design and OD estimation.
4. Understand the basic concepts of Connected and Autonomous vehicles (CAVs).
5. Apply the static network models to CAV planning and operations problems.
6. Use software to analyze various large-scale transportation networks and understand issues from real world practitioners.

Description:

This course provides an introduction to mathematical foundations of the analysis of transportation networks. The course will be divided into two main sections. Section 1 will introduce the basic foundations of network routing problems including user equilibrium (selfish routing) and system optimal games on networks. Various optimization based formulations, algorithms and extensions will be discussed. A particular emphasis will be on devising efficient algorithms and computation on city networks. Students will be expected to know how to design efficient algorithms for network analysis and implement them on various datasets. The second half of the course will tailor the network models learned in the first half to understanding the impacts of connected and autonomous vehicles (CAVs). This will be done by taking specific example problems such as autonomous intersection control, parking design, network design for CAVs etc. Recent research papers will form the basis for developing these models. Extensive use of intuitive arguments, counterintuitive phenomenon (paradoxes) and network structures will be utilized to illustrate many situations graphically. In addition, computing the solutions efficiently using various network algorithms will be discussed. The course is research based and students in addition to learning the concepts will extend the concepts to a research project to be finished within the semester.

Topics Covered:

Conceptual Foundation, User Equilibrium, Extensions of User Equilibrium, Network Models for Connected Vehicles

Prerequisites:

Undergraduate calculus, knowledge of probability, statistics, and linear algebra at the undergraduate level. Basic programming skills in languages such as Matlab or C++.

Applied / Theory:

Minimization problem, Graph theory User equilibrium problem, Solving algorithm, Stochastic user equilibrium, Network modeling for connected and autonomous vehicles