Agent-based discrete-event simulation model for no-notice natural disaster evacuation planning

Abstract

Over the past century, natural disasters have been often large-scale and overwhelming catastrophes resulting in irreparable and uncontrollable damage. In order to decrease the average evacuation travel time and reduce casualties, as well as improve the medical care for vulnerable populations in natural disasters, it is worthwhile to look at strategic disaster evacuation planning. In this paper, an agent-based discrete-event simulation (ABDES) modeling framework based on an embedded GIS module is developed for making no-notice natural disaster evacuation planning and improving the evacuation strategy periodically by using the proposed simulation model. Furthermore, in order to examine the applicability and extensibility of the proposed integrated GIS-based ABDES modeling framework, we conduct the experiments using several practical scenarios based on the city of San Francisco in California. The proposed evacuation simulation modeling framework can offer decision-makers significant insight on the traffic assignment and the configuration of resources such as evacuation vehicles and relief facilities that are required to complete successfully an evacuation against the no-notice large-scale disasters.