Transit-Based Evacuation Planning Considering Official Control of Stochastic Noncompliance Behavior
研究了大规模公共交通疏散中，考虑撤离者随机不服从行为及官方控制措施的规划问题，通过两阶段随机规划模型最小化总成本，并以重庆案例验证有效性。
This study tackles a large-scale transit-based emergency evacuation planning problem, which explicitly considers the often-neglected stochastic non-compliance behavior of evacuees and official non compliance control measures with respect to the assigned pick-up points. By describing stochastic non-compliance with a multinomial logit model, we propose two two-stage stochastic programming models, one that conducts official non-compliance control for all evacuees of certain demand points and a more general one that can implement various levels of non-compliance control at each demand point. These models minimize an expected total cost, which includes the pick-up point capacity expansion cost and the fixed and variable costs of non-compliance control before disasters, as well as the expected pick-up point capacity overflow cost and travel cost after disasters. Three models are examined for comparison, one considering non-compliance behavior without control, one assuming all evacuees comply with the pick-up point assignments, and one allowing victims of each demand point to be assigned to multiple pick-up points. With a case study on Chongqing, China, we demonstrate the effectiveness of our proposed model and obtain managerial insights for better practices of large-scale transit-based evacuations.