Destination And Route Choice Of Traffic Emergency Evacuation Based On Multi-Agent Simulation

As China's economic development and the accelerated process of urban modernization, the number, population and scale of cities are expanding, resulting in fire, explosion, public health accident, construction accidents and other man-made catastrophic events frequently. To effectively deal with the serious consequences of unconventional emergencies, it often need minimize the total travel time of evacuation, as well as reduce casualties and property losses. It makes the large-scale evacuation in emergency management research very important and difficult.In this paper, the highway traffic emergency evacuation management is set as a research background, Multi-Agent theory is set as a research-based method, random utility theory is used to describe traffic behavior, Jason and S-Paramics software are applied as a research platform. The behavior of evacuees is studied and the change of system evacuation clearing time is analyzed. All this can provide the emergency evacuation a theory support and experiment method.We study many knowledge from emergency evacuation theory, Multi-Agent theory, discrete choice theory and so on firstly. Then, a lot of work is done to analyze the behavior of evacuees in choosing emergency shelter -evacuation route and whether to change choice in the evacuation, at the same time, nested paired combinatorial Logit model and bi Logit model are established. This paper produce a questionnaire of imaginative evacuation in Changchun and use Internet-based SP survey as data investigation method, then Stata software and SPSS software are used to present the NPCL and BL model parameter calibration. The sensitivity analysis of parameters is done after that. Then the paper give a analysis on the inter-Agent interactions and information guiding method, including GPS car navigation systems, VMS variable message boards and radio, etc.This paper, combining the simulation software Jason which is suitable for Multi-Agent theory and the micro traffic simulation software S-Paramics, builds a special simulation environment. And then some evacuation management methods such as contra-flow lane, staged evacuation scheduling and elimination of intersection conflicts are used in the simulation process. A local road network in Changchun city is used, and then whether the information guidance is used and different combinations of information guidance strategy are discussed. Simulation results show that information can effectively reduce system evacuation clearing time, and GPS navigation system and VMS are both effective ways superior to radio in the roles of balancing network volumes and ease the increasing rate of the population in shelters.