A Hierarchical Path View Model for Path Finding in Intelligent Transportation Systems

  • Authors:

  • Yun-Wu Huang;Ning Jing;Elkea Rundensteiner

  • Affiliations:

  • University of Michigan, ywh@eecs.umich.edu;Chagsha Institute of Technology, ningjing@pdns.nudt.edu.cn;Worcester Polytechnic Insitute, rundenst@cs.wpi.edu

  • Venue:
  • Geoinformatica
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

Effective path finding has been identified as an importantrequirement for dynamic route guidance in Intelligent Transportation Systems(ITS). Path finding is most efficient if the all-pair (shortest) paths areprecomputed because path search requires only simple lookups of theprecomputed path views. Such an approach however incurs path viewmaintenance (computation and update) and storage costs which can beunrealistically high for large ITS networks. To lower these costs, wepropose a Hierarchical Path View Model (HPVM) that partitions an ITS roadmap, and then creates a hierarchical structure based on the road typeclassification. HPVM includes a map partition algorithm for creating thehierarchy, path view maintenance algorithms, and a heuristic hierarchicalpath finding algorithm that searches paths by traversing the hierarchy. HPVMcaptures the dynamicity of traffic change patterns better than the ITS pathfinding systems that use the hierarchicalA* approach because: (1)during path search, HPVM traverses the hierarchy by dynamically selectingthe connection points between two levels based on up-to-date traffic, and(2) HPVM can reroute the high-speed road traffic through local streets ifneeded. In this paper, we also present experimental results used tobenchmark HPVM and to compare HPVM with alternative ITS path findingapproaches, using both synthetic and real ITS maps that include a largeDetroit map ( 28,000 nodes). The results show that the HPVMincurs much lower costs in path view maintenance and storage than thenon-hierarchical path precomputation approach, and is more efficient in pathsearch than the traditional ITS path finding using A* orhierarchical A* algorithms.