Traffic modelling and simulation for regional multimodal evacuation analysis

Abstract

History has shown that evacuation planning has been based largely on lessons learned from prior events. While it yields incremental improvements, this approach also results in plans that tend to address failures 'of past evacuations', while leaving vulnerabilities to unforeseen conditions. Traffic simulation is a tool that can be used to evaluate evacuation plans and test contingencies not yet experienced or imagined. This paper discusses an effort to develop and test an agent-based travel demand and simulation model of the New Orleans regional multimodal evacuation plan. It is among the first to integrate a multimodal evacuation process into a regional traffic model with over a million vehicles, covering thousands of square miles, over two days and the first to use field recorded evacuation data for calibration and validation. The result of the model were used to quantitative and qualitative assess speed, volume, density, and congestion formation and recovery throughout the evacuation network.