Shortest hyperpaths in a multimodal hypergraph with real-time information on some transit lines

摘要

This paper presents an algorithm to find multimodal shortest hyperpaths in a transport system where transit arrival is random (i.e. random-arrival transit system), while real-time information on vehicle arrivals is only available for some lines and modes. When the algorithm is queried, there is a short horizon where real-time information is accurate, and past this the most reliable information for estimating the arrivals is to use a random distribution specific to the lines. This problem occurs frequently in emerging cities where the transit schedules are not maintained. A Combined Real-Time Hypergraph is constructed to model the multimodal transport system where all the public transport modes have random arrivals and real-time information on arrivals is available for some lines of some modes. In order to model the period of time when some lines have real-time information, the composition of the head nodes of hyperarcs changes over time. The algorithm is tested on a real-life transport system where we change the number of lines with available real-time information to assess the performance of the algorithm in different scenarios. We found that incrementing the number of lines with real-time information does not impact the performance.