Research On Routing Model Considering Risk-averse

The urban traffic flow system is a complex and uncertain system.In the actual travel process,it is difficult for the traveler to master effective and reliable real-time dynamic traffic information.At present,most researches on routing problem are focused on the shortest path optimization in the network targeting travel time or travel expenses.These studies assume that travelers have complete traffic information and are based on the theory of expected utility.The travel cost utility function is maximized,but it neglects the influence of the traveler's individual attributes.For any traveler,his/her travel is affected by his/her own age,gender,mentality,purpose and preference of travel,and externally provided traffic information.Each trip will have a different choice.Studies have shown that most travelers will exhibit risk aversion in the routing process.The public characteristics of risk-averse travelers are used to study the path choices for first-time and multiple-trip travel.Firstly,the traveler is affected by various subjective and objective factors in the travel process.The travel behavior in the uncertain traffic environment is a multi-stage decision-making process,and the neighboring nodes on the road segment foreground value and travel time reliability are established backwards.Recursive relations ensure real-time dynamic adjustment of the routing.With the theory of stochastic dominant is introduced,sorting the risk of the travel time of each alternative route,and considering the risk of delay from the other side,reflect the traveler's risk-averse behavior characteristics.The characteristics of decision-making behavior during multiple trips are analyzed and the travel behavior process is described.From the perspective of the traveler microcosmic,the individual travel process was modeled from the aspects of individual learning during the travel process,confidence in traffic information,and the initiation and termination of habitual selection behavior.According to the expected arrival time,the departure time selection problem is studied,including the departure time adjustment process and the arrival time threshold,and the risk avoidance traveler setting the threshold for the late arrival time.Finally,the routing simulation system is established via Repast platform and model is simulated.The analysis finds that the prospect theory is more effective than the expected utility theory in describing uncertainty.In addition,the routing behavior of travelers under different traffic information confidence degrees and different road network loads is studied,as well as how to adjust departure time.