A Hierarchical Route Planning Algorithm Based on Multi-level Topological Structure of Road Network
一种基于道路网络层次拓扑结构的分层路径规划算法

The efficiencies of planar shortest path algorithms deteriorate sharply with the expansion of the network size.However,this puzzle can be well resolved by hierarchical route planning algorithms which use "decompose and conquer" strategy to reduce search space essentially.Firstly,the paper studies the underlying data background of hierarchical algorithms,namely,the multi-level topological structure of road network,which covers road-class-specific level abstraction method region partition of road map data and intra-region hierarchical topological relationship model.Secondly,a hierarchical route planning algorithm is proposed for the optimal route calculation in location-based services(LBS).In detail,line distances between start nodes and goal nodes are applied to judge whether the node need to be switched to higher level.A modified heuristic A* algorithm is devised to search for the entrances to a higher level or the exits to a lower level.And a bidirectional strategy is adopted for intra-level optimal path computation.At last,some experiments using real road networks show that the proposed algorithm can largely improve the efficiency of route planning,especially for those cases with large-scale road network.