Resumen
A direct way of reducing the number of cars on the road is to dissuade individuals from exclusively using their car and instead integrate public transport into their daily routine. Planning multi-modal journeys is a complex task for which individuals often rely on decision support tools. However, offering individuals different journey options represents a significant algorithmic challenge. The failure to provide users with a set of journey options that differ considerably from one another in terms of the modes of transport employed is currently preventing the widespread uptake of multi-modal journey planning among the general public. In this paper, we introduce a dynamic programming algorithm that remedies this situation by modeling different transport networks as a graph that is then pruned by various graph-reduction pre-processing techniques. This approach enables us to offer a diverse set of efficient multi-modal solutions to users almost instantaneously. A computational study on three datasets corresponding to various real-world mobility networks with up to 30,000 vertices and 596,000 arcs demonstrates the effectiveness of the proposed algorithm.