Citation

Lee J, Kockelman KM. Transp. Res. Rec. 2021; 2675(9): 1398-1409.

Copyright

(Copyright © 2021, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.1177/03611981211007482

PMID

unavailable

Abstract

A scheduling algorithm is developed for optimal planning of large-scale, complex evacuations to minimize total delay plus travel time across residents. The algorithm is applied to the eight-county Houston-Galveston region and land use setting under the 2017 Hurricane Harvey scenario with multiple destinations. Autonomous vehicle (AV) use under central guidance is also tested, to demonstrate the evacuation time benefits of AVs. Higher share of AVs delivers more efficient evacuation performance, thanks to greater reliability on evacuation order compliance, lower headways, and higher road capacity. Furthermore, 100% AV use delivers lower overall evacuation costs and network clearance times and less uncertainty in travel times (via lower standard deviation in). Based on evaluations of different evacuation schedules, a 50% compressed evacuation time span resulted in longer travel times and network congestion. A 50% longer evacuation time span reduced residents' total travel time and network congestion, but increased the evacuation cost. As expected, evacuation efficiency falls when evacuees do not comply with evaucation schedules. Large shares of AVs will not be possible in the near future, so methods to enhance evacuees' compliance behavior (e.g., enforced and prioritized evacuation orders) should be considered until a meaningful level of AV technical maturity and penetration rate is available. This paper demonstrates the benefits of scheduled departure times, AV use, and evacuation order compliance, which help balance conflicting objectives during emergencies.

Language: en