Hyperpath-based algorithms for the transit equilibrium assignment problem

The transit equilibrium assignment problem (TEAP) aims to predict the distributions of passenger flows on lines or line segments in a transit network. Compared to the traffic assignment problem (TAP) for highway networks, the TEAP is much less studied, especially in terms of solution algorithms. This paper proposes two Newton-type hyperpath-based algorithms for a frequency-based TEAP formulation that considers both the congestion effect related to crowding and the queuing effect related to boarding. These newly developed algorithms, as well as two benchmark algorithms from the literature, are tested and compared on a number of transit networks, including two constructed using real-world data. The results show the proposed hyperpath-based algorithms significantly outperform the benchmark algorithms in large networks.