21 A Multi-Objective Optimization Model for Developing Emergency Evacuation Plans (PSAM-0466)

Evacuation of people from the hazardous region(s) is per se a way to reduce the ill effects of emergency events. The 2005 New Orleans's hurricane event showed that the efficiency of the evacuation system has great effect on the anti-disaster operations. The goal of evacuation planning is to maximize the performance of the emergency management system which is responsible for evacuating the residents in the threatened areas. The focus of this paper is on the development of a method to determine the appropriate management operations of evacuation planning. We extend the classical evacuation models that normally set for one target - the clearance time. It is found that even if the total clearance time is minimized, the migrating system may not be optimum because the congestion problem, the waiting time and the transportation network reliability are also the decision criterions. Generally, these targets conflict each other.

Our multi-objective optimization model is based on the analysis of the transportation network. Three objectives are used to represent the overall performance of the evacuation system. They are total evacuation time, average efficiency index and network robustness under disaster. The decision variables include two categories: traffic control actions and the schedule of escape from different zones. In this model, we initially merely consider the simplest human behavior under disaster: always choosing the shortest path to the safe places. The model can serve as a tool to devise efficient evacuation plans in advance.