Research On Travelers’ Target-oriented Stochastic Multi-attribute Route Utility Model And Its Application

Stochasticity is one of the typical characteristics of the practical traffic network,and this stochasticity could come from the demand-side and supply-side factors of the traffic network,travelers’ perception error and so on.Moreover,the empirical studies have shown that travelers’ route choice behavior can be affected by multiple attributes,e.g.,travel time,travel cost and travel time reliability;meanwhile,the empirical studies have also shown that there exist behavioral characteristics of travelers’ route choice behavior.Therefore,under this background,this thesis adopts the target-oriented analysis to develop travelers’ route choice model affected by stochastic multiple attributes.Based on this model,this thesis further analyzes the bi-attribute route choice model considering the impact of travel time and travel cost,and further proposes the bi-attribute user equilibrium on the stochastic traffic network.The detailed contents of this thesis are shown as follows.Stochasticity of the traffic network makes the deterministic attributes become stochastic,and thus,this thesis firstly proposes a target-oriented methodology to study travelers’ route choice behavior affected by stochastic multiple attributes on the traffic network,and proposes the target-oriented multi-attribute route utility model,which could belong to the descriptive paradigm.Meanwhile,the proposed model can be treated as a stochastic multi-attribute decision making model,which can provide implications for other disciplines,e.g.,the operations management.In the proposed model,travelers specify a target for each corresponding attribute,and the route utility depends on the targets that are achieved.There are two prominent characteristics of this model:(1)The correlation of the stochastic attributes is captured;(2)The imperfect complementarity and perfect complementarity relationship are proposed to model the target interaction based on the supermodularity of the utility function,and the meaning of the later one is that achieving more targets can bring extra utility to the travelers.Based on the above model,this thesis studies a bi-attribute(travel time and travel cost)route choice model on a stochastic traffic network with tolls,and further propose a target-oriented bi-attribute user equilibrium model,which is formulated as a variational inequality problem and solved on the Braess traffic network and Nguyen-Dupuis traffic network with the method of successive average.The target for route travel time is endogenously determined via travelers’ exogenously given on-time arrival probability,while the target for route travel cost is exogenously given,which can be determined by their income.This thesis theoretically proves and numerically demonstrates the impact of target on the equilibrium results(i.e.,the sensitivity analysis)in the application via four parameters,which are on-time arrival probability,target for route travel cost,time-to-cost utility ratio and complementarity ratio.The findings can provide new insights into traveler’s route choice behavior,which can further provide some implications for the policy design and implementation,especially on the congestion tolling.