Modeling And Simulation Of Path Selection Behavior Based On Type 2 Fuzzy For Emergency Evacuation

In recent years, due to the reasons such as natural and human factors, major disasters are causing around the world more frequently. Under such conditions, the influence of the incidents caused by large crowd evacuation problems are claimed highly academic attention. Comparing with the traffic flowing research in evacuation planning and management, evacuees behavior research is a weaker areas in the study of current traffic management. Considering the evacuees as the main body in the process of the emergency evacuation, differencing in predicting and choosing the evacuated time, route and destination will directly result and determine different regional traffic flow’s guidance performance.As evacuees’social psychology will change under different uncertainty levels and influent their decisions on choosing different evacuate routes, this paper will initially analyze the limitation of the traditional Logit model, i.e. it couldn’t clearly depict the evacuees’psychological behavior; and then, after analyzing the different uncertainty influencing factors, modeling evacuees behavior based on Type-II fuzzy theory and applied to the improved dynamic traffic flow distribution model. Finally examined and validated by Guangzhou Tianhe road network. Specific working contents are as follows:First of all, illustrating the limitation of traditional Logit model in the route choosing behavior by using examples, then after considering the risk perception behavior model under the influence of traffic conditions, introduce the weighted value of the path complex based on the obtainments from Type-II fuzzy method. Afterwards, author will improve the Logit model so that it could make a better description on evacuees’psychological behavior in the processing of emergency evacuation.Second part will focus on the comparison and analysis on the Type-I fuzzy theory and Type-II fuzzy theory and a detail discussion on the Type-II fuzzy system in determining the evacuate route, the design of path complex weighting model and two kinds of "two in, one out" Type-II fuzzy inference systems. All above are achieved through and based on the programming on Java platform.Finally, take the road network of Tianhe District in Guangzhou City as an actual case, and the EVAQ software (designed by Matlab) as simulation platform. The proposed evacuation behavior model based on Tpye-II fuzzy theory has been verified by this platform on analyze several indexes, such as evacuating time and persons, route network clearing time, etc..Simulation results show that higher level of traffic information may not improve the evacuation efficiency. Furthermore, evacuees prefer the routes with less intersections and short distance when they are close to destination.