Optimal deployment of autonomous buses into a transit service network
Autonomous vehicles empowered by emerging automation technologies are highly anticipated to be introduced into public transit service operations in the future mobility system. Considering the low acceptance rate of the new service with autonomous buses when it is initially put into practice, it is n...
Saved in:
Main Authors: | , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
2022
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/163462 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | Autonomous vehicles empowered by emerging automation technologies are highly anticipated to be introduced into public transit service operations in the future mobility system. Considering the low acceptance rate of the new service with autonomous buses when it is initially put into practice, it is not ideal to make a “one-off” deployment to replace all the service lines with autonomous bus services. Rather, the service operator is to determine an optimal plan for the deployment of autonomous buses onto different service lines in multiple stages. This paper proposes a multi-stage mathematical modeling framework to optimize the deployment strategy in which conventional buses are sequentially replaced by autonomous buses. More specifically, the model decides when (at which planning stage) and where (on which service line in the network) the deployment of autonomous buses should be conducted. Passengers’ acceptance attitudes towards autonomous buses are explicitly considered in their transit routing choices. To forecast the evolution of the passengers’ adoption rate of the autonomous bus service, a diffusion model is applied. The proposed multi-stage planning model framework, which is indeed a mixed-integer nonlinear program, is to determine the optimal deployment strategy that minimizes the total travel cost during the planning horizon. A two-phase solution method that combines a searching algorithm and a double projection method is proposed to solve the model. Finally, numerical studies are conducted to test the validity of the modeling framework and solution method. The impacts of passengers’ adoption rate and other parameters on the deployment strategy are illustrated. |
---|