Shortest-path and minimum-delay algorithms in networks with time-dependent edge-length
Journal of the ACM (JACM)
Dijkstra's algorithm on-line: an empirical case study from public railroad transport
Journal of Experimental Algorithmics (JEA)
Pareto Shortest Paths is Often Feasible in Practice
WAE '01 Proceedings of the 5th International Workshop on Algorithm Engineering
Efficient models for timetable information in public transportation systems
Journal of Experimental Algorithmics (JEA)
ESA '08 Proceedings of the 16th annual European symposium on Algorithms
Engineering Route Planning Algorithms
Algorithmics of Large and Complex Networks
Car or Public Transport--Two Worlds
Efficient Algorithms
Robust and Online Large-Scale Optimization
Contraction hierarchies: faster and simpler hierarchical routing in road networks
WEA'08 Proceedings of the 7th international conference on Experimental algorithms
Multi-criteria shortest paths in time-dependent train networks
WEA'08 Proceedings of the 7th international conference on Experimental algorithms
Time-dependent contraction hierarchies and approximation
SEA'10 Proceedings of the 9th international conference on Experimental Algorithms
Fast routing in very large public transportation networks using transfer patterns
ESA'10 Proceedings of the 18th annual European conference on Algorithms: Part I
Parallel computation of best connections in public transportation networks
Journal of Experimental Algorithmics (JEA)
Exact Routing in Large Road Networks Using Contraction Hierarchies
Transportation Science
| Hi-index | 0.00 |
We contribute a fast routing algorithm for timetable networks with realistic transfer times. In this setting, our algorithm is the first one that successfully applies precomputation based on node contraction: gradually removing nodes from the graph and adding shortcuts to preserve shortest paths. This reduces query times to 0.5 ms with preprocessing times below 4 minutes on all tested instances, even on continental networks with 30 000 stations. We achieve this by an improved contraction algorithm and by using a station graph model. Every node in our graph has a one-to-one correspondence to a station and every edge has an assigned collection of connections. Also, our graph model does not require parallel edges.