Journey Levels in Strategy-Based Transit Assignment: Modeling Integrated Transit Fares and More

The strategy-based transit assignment has been used extensively and its properties have been well understood since it was first developed. It can be used to model trip-additive generalized user costs, such as those incurred when waiting, boarding, walking, or traveling in-vehicle during a transit journey that can include transit fares. However, the solution algorithm is Markovian and does not directly incorporate the journey-dependent state, such as the sequence of modes used to arrive at a node, which is necessary to model fares in integrated transit fare schemes. In practice, network constructions have been used to add state information to the strategies method; doing so adds complexity to and increases the size of the transit network model. Elaborate network constructions, which duplicate some or all of the transit network and add new transition links, may be necessary for more complicated fare schemes. An extension is proposed to the strategy-based transit assignment in which user costs can be made to vary along a trip according to a journey-dependent state defined by a set of transit journey levels. The journey-level enhanced strategy transit assignment algorithm allows explicit modeling of networkwide integrated fare schemes without network construction and provides a compact and flexible expression for the state transitions and corresponding generalized costs involved. The corresponding algorithm maintains the computational efficiency of the strategies method. Examples are provided together with computation results on small and large networks.