A dynamic line generation and vehicle scheduling method for airport bus line based on multi-source big travel data

被引:0
作者
Haitao Yu
Weifeng Lv
Hangou Liu
Xiaoning Fu
Randong Xiao
机构
[1] Beihang University,School of Computer Science and Engineering
[2] Beijing Transportation Information Center,College of Software Engineering
[3] Beijing Key Laboratory for Comprehensive Traffic Operation Monitoring and Service,undefined
[4] Beihang University,undefined
来源
Soft Computing | 2020年 / 24卷
关键词
Airport bus line; Demand responsive transit; Candidate stations; Station grouping; Line generation and vehicle scheduling; Public transportation demand level;
D O I
暂无
中图分类号
学科分类号
摘要
Airport bus is an important public transportation mode for large international airport. To improve the bus station coverage, passenger demand compatibility and the scheduling flexibility of Beijing International Airport bus line, a dynamic line generation and vehicle scheduling method is proposed in this paper. Firstly, based on multi-source big data from the airport (including data from taxi, ride-hailing service, subway, regular bus, airport bus, etc.), we accurately extract candidate stations, which are very popular with passengers and convenient for parking and transfer, through public transportation demand level calculation, iterative clustering and POI matching. Then, the candidate stations need to be partitioned appropriately by selecting suitable features and calculating the similarity of candidate stations, so as to make the stations within each group a moderate size and have a consistent spatial orientation. Finally, a line generation and vehicle scheduling algorithm, which is compatible with multi-vehicle, high success rate of ride-sharing matching and low cost, is designed to realize accurate and rapid operation scheduling within each group according to the situation of passengers booking tickets. We have carried out experiments in Wangjing and Yayuncun, and the results show that our method can satisfy passenger demand fast and accurately.
引用
收藏
页码:6329 / 6344
页数:15
相关论文
共 33 条
[1]  
Dessouky M(1994)Bus dispatching at timed transfer transit stations using bus tracking technology Transp Res Part C Emerg Technol 7 187-208
[2]  
Hall R(2004)Procedures for planning multi-leg journeys with fixed-route and demand-responsive passenger transport services Transp Res Part C Emerg Technol 12 33-55
[3]  
Nowroozi A(2002)Multi-modal and demand-responsive passenger transport systems: a modelling framework with embedded control systems Transp Res Part A 36 167-188
[4]  
Horn MET(2011)Dynamic ride-sharing: a simulation study in metro Atlanta Transp Res Part B Methodol 45 1450-1464
[5]  
Horn MET(2011)Massive scale-out of expensive continuous queries Proc Vldb Endow 4 1181-1188
[6]  
Agatz N(2000)FDBSCAN: a fast DBSCAN algorithm J Softw 11 32-34
[7]  
Erera A(2008)Efficient generation of simple polygons for characterizing the shape of a set of points in the plane Pattern Recognit 41 3224-3236
[8]  
Savelsbergh MWP(2012)Fast affinity propagation clustering: a multilevel approach Pattern Recognit. 45 474-486
[9]  
Zeitler E(2013)Semi-supervised traffic identification based on affinity propagation ACTA 39 1100-1109
[10]  
Risch T(2008)Semi-Supervised clustering based on affinity propagation algorithm J Softw 19 2803-2813