Research Of Public Transit Route Choice Methods Based On The Instant Interaction Of Information

As an important part of urban transportation,public transportation has obvious advantages,which can play a good role in easing traffic congestion and improving urban traffic conditions.However,the passenger volume of ground public transport in some Chinese cities,especially in megacities,has been decreasing year by year.Besides,the sharing rate of public transport in Chinese cities is still at a low level.Problems such as long waiting time,slow driving speed,poor in-vehicle environment and inconvenient transfer are still prominent.Under the background of Internet era,the rapid development of communication technology,especially the wide application of smartphone,provides an opportunity for the development of public transportation.Based on the information instant interaction between passengers and intelligent terminals such as smartphones,this paper divides the real-time information in the instant interaction process into three levels according to the order of appearance and level of richness,which are 1)the first level of information which is widely applied in smartphone application for bus traveling now includes passenger preferences,vehicle arrival and routes in the transit network;2)the second-level information,which is applied in the emerging smartphone application for bus traveling recently,includes the first-level information as well as in-vehicle congestion information;3)the third level of information,which has appeared in no smartphone application until now but will be applied in the future intelligent terminals,includes the second level of information as well as passengers’ previous travel experience.After that,passengers’ route selection behaviors and transit assignment results under information of different levels are analyzed.The main work and innovation points of this paper are summarized as follows:(1)According to the quantity of real-time information received by passengers unfamiliar with their traveling routes,three route choice methods considering respective travel scenarios are proposed,which are as follows:1)without considering instant interaction and real-time information,a first-come-first-serve route choice method is designed for the scenario;2)with considering no instant interaction and only receiving vehicle arrival information from bus electronic station sign,a route choice method using the discrete choice model is designed by considering factors such as waiting time at stops;3)under the condition of the first level information instant interaction,according to the scenario that passengers send their travel preferences through intelligent terminal to obtain multi-dimensional travel information,a route choice method is designed considering passengers’ travel preferences and instant interaction in which a three-phase model is constructed which involves the generation,choice and adjustment of travel route strategy.The stage of travel route generation determines the shortest route set using the deep first and forward and back search method.The stage of route choice constructs a utility model considering five traveling attributes including waiting time,walking time,in-vehicle time,number of transfers and traveling fare.The stage of route adjustment is divided into two periods based on whether a passenger has boarded a bus.The purpose of the adjustment strategy is to ensure that passengers have a complementary travel strategy beyond the tolerance level when waiting too long before boarding or feeling too crowded after boarding.The three travel route choice methods are tested on the simulation platform and the vehicle capacity limitation is also considered.The results show that under the condition of information instant interaction,the travel cost of unit passengers can be reduced by 1.01 yuan and 2.6 yuan compared with the other two methods.(2)Under condition of the second level information instant interaction,a within-day route choice model is built by receiving the in-vehicle congestion information,as well as passengers’ preference information in view of a scene that passengers are unfamiliar with traveling routes searching from intelligent terminal,which includes the process of choosing the initial travel route before boarding and the transfer travel route after boarding.Besides,an information attenuation model is also utilized considering the character of information attenuation which establishes the relationship of passengers’trusting the receiving information and the degree of information attenuation.Additionally,four predicting in-vehicle congestion information models involving the request-based prediction model,historical data-based prediction model,probability distribution-based prediction model and multi-methods connection-based information prediction model are also proposed based on the resource,accuracy and promptness of data.The results show that under the condition of providing in-vehicle congestion information,the available number of seats of passengers can be significantly increased,the vehicle congestion rate and fail-to-board passengers can be reduced,and thus the in-vehicle environment can be effectively improved.In addition,the condition of providing in-vehicle congestion information can effectively balance the vehicle passenger flow in the transit network,so as to optimize the allocation of resources.(3)Under condition of the third level information instant interaction,a day-to-day route choice model considering passengers’ travel experience and preference as well as the in-vehicle congestion information is proposed.In this model,wait time,walking time,in-vehicle time,number of transfers,travel fare,congestion cost and other travel attributes are considered to build a route choice model based on the improved experience-weighted attraction self-learning model.In addition,in order to depict the dual effects of rational knowledge and travel habits on passengers in the process of choosing route,a rational-experience model is proposed,which considers the probability of rational choice on behalf of rational knowledge and the probability of experience choice on behalf of travel habits,which is more practical to describe the passengers’ travel process.