Accelerating the k-shortest paths computation in multimodal transportation networks

Intermodality in transportation systems is fast becoming a research topic of great interest. The computation of multiple paths in a multimodal network is desirable to identify efficient routes based on user preferences. This paper describes the clustering technique, which improves the performance of conventional k-shortest paths computations in multimodal transportation networks by first transforming the network into an acyclic representation through identification of cycles and clustering them into hypothetical nodes. The generalized Floyd algorithm is then applied on these clusters to compute the k-shortest paths. Simulation results show that the proposed technique significantly improves the performance of the conventional algorithm, particularly when the required number of k-shortest paths increases.