Large-scale transit service network design under continuous heterogeneous demand

transit network design;hybrid network;heterogeneous demand

Smith, Ryan ; Ouyang ; Yanfeng

关键词: transit network design;    hybrid network;    heterogeneous demand;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/49736/Ryan_Smith.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

The design of public transit systems typically relies on discrete or continuous models todetermine route layout and service frequency. To avoid computational complexity associatewith discrete models, continuous models are proposed to formulate the problem in terms ofa few key continuous variables. One drawback of many continuous models is that they typicallyassume a uniform distribution of passenger trips, however we know that trip origins anddestinations are characteristically not uniform. The purpose of thesis is to (i) investigate howthe design and operations of Daganzo (2010b) hybrid transit system is effected by spatiallyheterogeneous demand; and (ii) how large is and where to locate a transit network undera spatially heterogeneous demand. The spatially heterogeneity of the passenger demand iscaptured in the model by transforming a continuous passenger demand density function intozone-to-zone passenger demand. For (i), the optimal hybrid transit structure and operationsis analyzed for two distinct spatial demand distributions each with a low and a highpassenger demand scenario. The results show the effects on agency and user cost metricswhen passenger demand is increased across the region and concentrated (i.e, the origins anddestinations of passengers trips are grouped more tightly together in the center of the city).For (ii), automobiles are introduced as an alternative mode of travel and each person whochoose this mode incurs a cost associated with not taking transit. This is formulated andpeople with an origin and/or a destination outside the transit service region are assumed tonot take transit. For each transit mode a sensitivity analysis is performed on the penaltyfor not taking transit. The results indicate that the optimal size of the transit service region(i.e., the number of passengers served by the transit system) is dependent on the not takingtransit penalty and the cost and operational characteristics of the transit mode. Finally,multiple future research ideas are presented to illustrate some of the other possibilities thatmay further enhance transit network design.