Research On Agent-based Travel Mode Choice Behavior Of Autonomous Vehicle

As one of the transformative technologies in the transportation industry,autonomous driving technology will have an important impact on road networks,congestion,safety and the environment.In particular,due to the increase in travel modes and the consequent changes in the supply of travel mobility,the travelers’ mode choice behavior will change.This research mainly considers the traveler’s mode choice behavior among private autonomous cars(PAVs),shared autonomous cars(SAVs)and traditional cars.Existing researches on autonomous driving travel mode choice behavior mainly adopts two methods: discrete choice model and agent-based model.The factors that mainly affect the travelers’ mode choice behavior are socio-demographic attribute,family attributes,travel mode attributes and traveler’s preferences.It is difficult to use the discrete choice model to consider the influence of direct interaction between travelers on the mode choice behavior,and the influence of traveler’s travel experience on mode choice can be considered in the agent-based models,this study will use agentbased model to study the travelers’ mode choice behavior.At the same time,this paper take travel experience and social interaction into consideration.Then the size of parameters of travel experience and social interaction would also be took into consideration.In order to make sure to what extent the influence of observable factors on traveler’s choice would be,the questionnaire is used to understand the traveler’s choice behavior,and then the conditional logit model is used to calibrate the selection results.The results indicate the traveler’s socio-demographic attributes,family attributes and travel mode attributes will affect traveler’s mode choice behavior.Then the results of conditional logit model are substituted into the agent-based model,and the comprehensive influence of travel experience and social interaction is considered to obtain the final selection result.The simulation results show that both travel experience and social interactions have influences on the travelers’ mode choice.The travel experience would make the traveler’s mode choice result to be consistent with the result without preference.Social interaction will make the travelers to choose the same travel mode with those around them.At the same time,social interaction will make travelers prefer to choose the mode with a middle utility value.This research studies the travelers’ mode choice behavior with agent-based simulation,and tries to explain the influence mechanism of travel experience and social interaction on travelers’ mode choice behavior,and provides certain data for vehicle companies to find potential autonomous vehicles users and provides traffic manager some assistance in policy making of autonomous vehicles.