Random Regret Route Choice Behavior Based Evacuation Traffic Equilibrium Analysis And Its Application

Emergency evacuation is a special type of travel activity motivated by active hedging. During the movement from dangerous area to safe place, transportation system plays an irreplaceable role during the process of emergency evacuation. Currently, many cities have made emergency transportation plans for disaster events in order to reduce property damage and human casualties, and promote the disaster mitigation ability.In recent years, it has been acknowledged that anticipated regret is one of the most important determinants of choices. Regret arises when a considered alternative is outperformed by another alternative, especially in emergency con text. Facing the road network under damage risk affected from disasters, evacuees often exhibit different behaviors from daily travel. They are not only pure utility-maximization pursuers, also try to avoid the negtive regret emotion produced by the chosen route alternative. Once evacuees’ attitude and preference are ignored, it is difficult to make a scientific evacuation transportation plan due to cognition biases on a macro-level evacuation network. Therefore, combining random utility theory and random regret theory, the thesis follows the research mainline of travel behavior choice theory, route choice behavioral analysis, evacuation traffic flow equilibrium analysis, and evacuation network optimal design.The development course and empirical analysis on random regret minimization model were reviewed. In contrast to random utility maximization model, the necessity of regret decision rule based application on evacuation behavior was discussed, as the theoretical basis of further evacuation route analysis. Evacuation travel behivors under typhoon disaster based on stated preference survey were investigated. Evacuation route choice data were obtained on different attribute factors including route distance, accessibility, perceived damage risk and service level as database for further route choice analysis.On a microcosmic individual level, utility maximization rule and regret minimization rule were jointly considered on evacuation route choice problem. Based on random regret theory, a modified generalized regret based evacuation route choice behavioral model was formulated by introducing regret weights. Validated by stated preference data, estimated results of parameters were compared with that of random utility maxmization model and random regret minimization model. The influence of route attributes and evacuees’ information on route choice were discussed, and evacuees’ regret averse behavior on route choice was illustrated for further evacuation traffic flow analysis.On a macro roadnetwork level, the problem of evacuation traffic flow equilibrium was studied based on evacuees’ regret averse route choice behavior. According to stochastic user equilibrium theroy, a generalized regret paradigm based stochastic user equilibrium condition was defined. A regret averse evacuation route choice based stochastic user equilibrium model was constructed considering route overlap effect. Then, it can be extended to be a multi-class stochastic user equilibrium model for mutiple types of evacuees. Method of successive average was designed to solve the problem. The influence of relative importance of attribute, demand scale, and assignment parameters on evacuation traffic equilibrium state were analyzed.Take emergency manager and evacuees as decision subjects, evacuation network optimal design was proposed at evacuation traffic equilibrium state based on regret averse route choice. The critical links were recognized through evaluating evacuation performance of road network, and then a bilevel planning model for continuous design of evacuation road network was formulated to describe a Stackelberg game between emergency manager and evacuees. The Genetic Algorithm was applied to solve the problem. A reasonable allocation on limited emerency resources can be finished to protect network, and achieve a optimal road network evacuation performance. Take Double Happiness Island in Fujian as a case, the network design results were analyzed.In this thesis, the research of emergency evacuation transportation problem was conducted from regret psychology view. A more real and comprehensive cognition on evacuation process was completed from the micro individual behavior to the macroscopic road network flow. A optimal design method of evacuation road network has been given, and this thesis can provide scientific basis and theoretical support for an improvement on emergency response ability in a city or area.