Adjustment And Scheduling Optimization For Long Distance Bus Lines Based On Public Transit Big Data

The pure electrification of public transport is an important part of the national "Blue Sky Defense" and "Transit-oriented Cities" strategy.In the process of vigorously promoting the pure electrification of public transport in big cities of China,it has brought new challenges to the operation of public transport,such as the adjustment of bus lines and the layout of charging piles.There are a certain proportion of ultra-long bus lines in the bus network of each big city.In order to adapt to the development trend of pure electrification,it is urgent to optimize and adjust them.Therefore,this paper takes an ultra-long bus line as an example to realize the adjustment and scheduling optimization of the bus line by collecting and analyzing the corresponding bus big data,so as to meet the travel needs of passengers.First of all,according to a month of actual bus card swiping and entering and exiting station data,the passenger flow characteristics of this line and the rule of travel time are analyzed in detail.The first is to generate passenger flow OD matrix of uplink and downlink of the line by combining passenger card swiping rate.Second,according to the bus operation frequency data of the line,K-means method is adopted to divide the operation peak and off-peak time of the line.The third is to analyze the data of the bus entering and exiting the station,and get the distribution rule of bus dwell time at the stations and travel time between stations.Secondly,based on the minimum impact of passengers on the original route,under the constraints of the bus route length and the actual bus station location,various route adjustment schemes are obtained by calculating the coverage rate of passenger flow OD pair.On this basis,aiming at the minimum waiting time for passengers to transfer,eM-Plant was used to conduct simulation evaluation on the adjustment scheme,and the coverage rate of OD pair of passenger flow was integrated to obtain the final adjustment scheme of ultra-long line.Finally,the bus scheduling scheme among the adjusted sub-lines is co-optimized in this paper.A stochastic correlation opportunity programming model for passengerwaiting time cost,enterprise operating income and passenger transfer waiting time cost was established,and the model was solved by combining neural network and genetic algorithm to determine the bidirectional departure interval of the adjusted sub-line.eM-Plant simulation software is used for simulation verification.The results show that the average waiting time of line passengers is significantly reduced when the revenue of bus enterprises is slightly reduced and the waiting time of passengers to transfer meets the expectation.