Finding a Real Passenger Path in a Complex Transit Network Using a Smart Card Record

摘要

Korea has a very elaborate transit trip data, and smart card record. The original trip stop and destination stop can be accurately recorded when using public transportation in Korea, especially within the Seoul Metropolitan Area (SMA). This data can be used to calibrate path choices based on the stochastic transit assignment model, which can describe passenger behaviour more precisely. However, the model has to decide the alternative path set in a huge network. This paper reviews search methods of multiple reasonable paths to implement multimodal public transportation assignment models with the network of Seoul. With a large scale multimodal public transportation network such as Seoul, the computation time of a path finding algorithm is key, and the resulting paths should reflect the route choice behavior of the public transportation passengers. Search methods of alternative paths are divided into the removing path method and the deviation path method. In order to apply path finding algorithms to public transportation networks, transfer and loop constraints must be included to be able to reflect real behavior. To validate the availability of the algorithms, experiments conducted with Seoul metropolitan public multimodal transportation network consisted of 41,950 nodes and 171,176 links by using the deviation path method, which is suitable for a large-scale network.