Impact of Autonomous Vehicles on Household Activity and Travel Scheduling: An Integrated Dynamic Network Modeling Approach

Xu, Xiang

doi:https://doi.org/10.21985/n2-b8mb-p019

Work

Impact of Autonomous Vehicles on Household Activity and Travel Scheduling: An Integrated Dynamic Network Modeling Approach

Public

Download PDF

Download All Files (.zip)

Autonomous vehicles (AVs) have gained more and more attention from researchers, manufacturers, and transportation experts. When AVs become available to the public, they are expected to dramatically alter the transportation system as we know it, posing critical questions such as: How will the road network react to this change? How will AVs alter network performance? How will AVs influence traveler behavior and mode choice? Evaluating and predicting AVs’ impact, especially on household activity and travel schedule, calls for a holistic approach. This dissertation proposes an activity-based modeling and dynamic traffic assignment (ABM-DTA) integration framework with a privately operated autonomous vehicle (POAV) optimizer. Note that the privately operated AVs are privately controlled by the household, rather than by a fleet manager or central authority. And they are driven by machines instead of human. The proposed framework is applied to an integrated transportation and transit network for estimating the performance of the solution to the POAV optimization problem and evaluating POAVs’ impact on the whole transportation system. The dissertation begins with reviewing existing literature for AV-related problems. Next, the dissertation presents a first-order approach integrated with ABM-DTA framework to model the impact of AVs on household travel and activity schedules. Considering shared rides among household members, mode choices, re-planning of departure times, and the rescheduling of activity sequences, a mixed integer programming and several extensions are presented. The proposed approach is tested for different models at the household level with different household size and at the network level under different scenarios. Lastly, a fixed-point problem is proposed to ensure equilibrium is achieved among framework component by minimizing the gap between the traveler’s expected travel time and their experienced travel times simulated by the DTA tool.

Creator