Loading...

Simultaneous Traffic Assignment Model for Autonomous Vehicles and Regular Vehicles with Different Route Choice Criteria in Transportation Network

Mousavi, Roozbeh | 2020

563 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53207 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Zokaei Ashtiani, Hedayat
  7. Abstract:
  8. With the arrival of autonomous vehicles in the fleet of the transportation network and the change of some travel parameters such as travel time, mode choice models, as well as route choice models, will also change. These types of vehicles will be able to move less distance from the front vehicle compared to regular vehicles. Also, due to the exchange of information between the vehicle and the infrastructure, we will see an increase in the capacity of the transportation network and its safety. The period of complete transition from regular vehicles to self-driving vehicles will be relatively important and therefore it is necessary to see these two phenomena together. It is assumed that autonomous vehicles will act as a user equilibrium in the route choice stage due to receiving complete information about the network status. In regular vehicles, due to the driver's lack of information from the transportation network, it can not be expected that these types of vehicles will act according to the user equilibrium approach. One of the models that have been used for regular vehicles is the cross nested-logit model. These types of models, due to their distinctive structure, avoid the problem of overlapping traffic flow in the arcs, and are a relatively more realistic model for modeling the network of regular vehicles. In this study, each group of vehicles is classified in a separate class and a multiclass traffic assignment model is formulated for them. To solve the problem of multiclass traffic assignment, four solution methods have been proposed and their performance in Wang, Nguyen, Hearn and Sioux Falls networks have been compared with each other. All of these methods require primary paths to get started. The path generation for each origin-destination is performed using the Yen's K-shortest paths algorithm. Finally, the assignment- path generation algorithm is presented as the proposed solution of this study and the results of its implementation for Hearn and Sioux Falls networks are given
  9. Keywords:
  10. Autonomous Vehicles (AVs) ; Mode Choice Model ; Cross Nested Logit Model ; Multiclass Traffic Assignment ; Multi Modal Traffic Assignment

 Digital Object List

 Bookmark

  • attachment1.pdf
    • Bookmarks