Research On Public Transportation Scheduling Optimization Based On Deep Fusion Of Passenger Transfer Data

With the rapid development of economy and society as well as the growth of urban population,residents' travel volume keeps increasing,and the construction speed of transportation infrastructure is far behind the growth rate of traffic demand.Therefore,traffic congestion is increasingly prominent,which directly affects people's travel quality.Improving the operation efficiency and service quality of the public transportation system is a key measure to alleviate traffic congestion.The operation of public transportation system is a complicated problem,including the planning of bus network and the scheduling of drivers.Among them,scientific allocation of bus schedule is the key to public transportation scheduling,and is the most important and complex problem in system operation.Under the existing bus network structure,the transfer behavior of bus passengers is common.In order to improve the transfer efficiency of passengers in the bus network,this paper integrates multi-source passenger transfer data,analyzes passenger transfer behavior,and based on this,establishes a collaborative optimization model of bus schedule,and designs a heuristic algorithm to solve the model.First of all,based on multi-source data such as IC card data of bus passengers,GPS data of bus vehicles and GIS data of bus stops,the algorithm of passenger transfer behavior identification is constructed from two dimensions of space-time,and passenger transfer behavior is analyzed.In the time dimension,the time interval before and after the passenger swiping the card is divided into two parts: the journey time from the station to the station and the waiting time.Based on the transfer time threshold,the pre-transfer time of the passenger and the time of walking to the transfer station are calculated.In the spatial dimension,the spatial location of the bus station and the shortest transfer distance are used to judge the passenger's arrival and transfer station.Secondly,a collaborative optimization model for the schedule of double-layer and multi-objective bus lines is built.The upper goal of the model is to maximize passenger transfer efficiency,and the lower goal is to minimize passenger transfer time.The maximization of passenger transfer efficiency is based on the transport capacity of the bus,enabling as many passengers as possible to transfer to the bus within the transfer time window;The shortest transfer time for passengers is the shortest total transfer time for all passengers in the bus transfer network.According to the complexity of model calculation,an enumeration algorithm and a genetic algorithm with elite strategy(NSGA-?)based on non-dominant ordering are designed to solve the problem.Finally,two cases are selected to verify the effectiveness of the algorithm and model respectively.In terms of validation algorithm,through the case of a transit network,adopt the enumeration algorithm and with the elite strategy based on non dominated sorting genetic algorithm(NSGA-?),and compared the results of two algorithms and analysis,the results show that based on the non dominated sorting genetic algorithms with elite strategy compared with enumeration algorithm can get high quality in a reasonable amount of time the pareto solutions.In terms of model validation,through the second case of transit network,the collaborative scheduling model proposed in this paper compared with previous research collaborative scheduling model and evaluation,the results showed that the collaborative scheduling model in this paper the optimized passenger average transfer time compared with previous research model optimized percentage increase 10% average passenger transfer time,in the perspective of passenger travel cost,in this paper,the collaborative scheduling model is superior to the previous research the coordinated scheduling model.Through the analysis and research of the above two cases,it is proved that the collaborative scheduling model and solution algorithm proposed in this paper have a good application effect in practical engineering,which can improve the service level of public transportation network,attract travelers to choose public transportation as a way of travel,and thereby alleviate urban traffic congestion.