作者单位: |
1Ph.D. Candidate, School of Transportation and Logistics, Southwest Jiaotong Univ., 111 the Second Ring Rd. North, Chengdu, Sichuan 610031, China.
2Associate Professor, Ingram School of Engineering, Texas State Univ., San Marcos, TX 78666 (corresponding author). Email: f_w34@txstate.edu 3Research Associate, Institute for Multimodal Transportation, Jackson State Univ., 1230 Raymond Rd. 900, Jackson, MS 39204.
4Professor, School of Transportation and Logistics, Southwest Jiaotong Univ., 111 the Second Ring Rd. North, Chengdu, Sichuan 610031, China. |
摘要: |
A hurricane can cause a lot of damage to highway networks and significantly affect the evacuation trip operations on the roads. A good understanding of network vulnerability can improve preparedness for an emergency evacuation. This paper presents a game theory– based approach to the analysis of network vulnerability under a hurricane evacuation. A zero-sum game was constructed between a router, which seeks the minimum-cost paths for the evacuation trips, and a tester, which maximizes the travel cost by disturbing the links. The distribution of the evacuation demand was elastic because the probability of selecting an evacuation destination was determined by the path risk and travel cost. In addition, the congestion effect was also considered. A solution strategy based on the method of successive averages (MSA) was adopted. Over a sample network, the proposed method was compared with three other methods for the network vulnerability analyses. The method was applied to the vulnerability analysis of a realistic evacuation network in the Mississippi Gulf Coast area. |