Research On Multi-mode Long-distance Public Transport Scheme Based On Dynamic Random Network Of Passenger Preference

With the increasingly frequent economic and cultural exchanges across the country,in order to meet people's growing demand for long-distance travel,the public transport system has become increasingly rich and complex.Since large long-distance public transport networks usually involve many different routes and public transport modes,it is of great importance to provide travelers with personalized optimal route selection based on random route characteristics and provide theoretical support and application reference for convenient passenger travel.Therefore,aimed at the urgent needs and inadequacy of the public traffic path planning research on multi-mode public transportation route planning based on dynamic stochastic network of passenger preference,based on passenger preference and random route characteristics,this paper introduced the related theories of dynamic random network and prospect theory,and focused on the construction and solution of multi-mode public transportation path planning model based on the dynamic random network of passenger preference.The model was applied to the empirical study of the long distance public transportation path planning system and verified by a specific example.Firstly,the cohesion of public transport modes is studied by taking the most common long-distance bus,passenger train and civil airliner as the research objects.;Secondly,this paper discusses the construction of the multi-mode long-distance public transport travel scheme model,including the basic schedule problem and the multi-mode schedule problem,which lays a foundation for the use of topological sorting algorithm to solve the shortest path problem;Again,according to Dijkstra algorithm to obtain the preliminary recommended path,and drew lessons from the solutions of the research scholars at home and abroad to solve the dynamic stochastic shortest path by dividing network nodes up to several areas.The paper used the Dijkstra algorithm for determining the redundant areas and nodes,and then deleted them,using genetic algorithm to obtain many kinds of transport mode recommended path;Finally,aiming at the problem of the uncertainty of traffic conditions and the combination of multi-mode when travelers face the choice of multi-mode public transport mode,a multi-mode public transport mode selection model based on prospect theory is proposed to evaluate and select multiple paths obtained by genetic algorithm.When the genetic algorithm obtained the recommended path of the dynamic random network,the model size could be reduced to about 30% and the arithmetic speed could be increased by 10 times after redundant nodes were removed according to the strategy in the paper.Moreover,the expected value of recommended path conversion cost obtained by genetic algorithm was less than 20%,and the path size of prospect theory analysis was greatly reduced.In prospect theory analysis part,the model overcome faults of the traditional utility maximum expected utility theory assumption and traveler completely rational assumption,and to quantify the different travelers' travel cost,eliminating the deviation of traditional method,which took travel cost and travel time as a measure of the travel cost,and made the model more close to the reality of travel.