An Efficient Shortest Distance Decomposition Algorithm for Large-Scale Transportation Network Problems

For numerous large-scale engineering and science problems, domain decomposition (DD) has generally been accepted by research communities as among the most attractive methods to obtain solutions efficiently. As a pre-requisite for the DD solution process, a large domain must be partitioned into several smaller sub-domains, with the key to success (of any DD partitioning algorithm) being the number of system boundary nodes. The lower this number, the more efficient the sub-domains can be processed in parallel. Although various transportation researchers have hinted at the use of DD, for example, in decentralized traffic management, it is always assumed the partition is provided. This paper presents a simple, efficient and effective algorithm to decompose a transportation network into a predefined number of inter-connected sub-domains such that the number of system boundary nodes is small (first priority) and the number of nodes in each sub-domain is similar (second priority). This algorithm was based on a simplified version of the Polynomial (partitioned) Label Correcting Algorithm (P-LCA), rather than the classical (Bellman-Ford) LCA, coupled with simple heuristic rules. To assess the effectiveness (in terms of minimizing the number of system boundary nodes) of the proposed Shortest Distance Decomposition Algorithm (SDDA), it is compared with METIS, which is currently the most widely used graph partitioning algorithm world-wide. Using large-scale, real-world transportation test networks, it was found the SDDA is significantly better than METIS; the SDDA outperformed METIS in 23 of 27 examples, and on average provided (approximately) 46% fewer boundary nodes for large-scale examples.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ADB30(7) Paper Review Group #3. Alternate title: Efficient Shortest-Distance Decomposition Algorithm for Large-Scale Transportation Network Problems.
• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Johnson III, Paul W
  • Nguyen, Duc
  • Ng, ManWo
• Conference:
  • Transportation Research Board 93rd Annual Meeting
  • Location: Washington DC
  • Date: 2014-1-12 to 2014-1-16
• Date: 2014

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 18p
• Monograph Title: TRB 93rd Annual Meeting Compendium of Papers

Subject/Index Terms

• TRT Terms: Heuristic methods; Network nodes; Shortest path algorithms
• Uncontrolled Terms: Domain decomposition; Network decomposition; Transportation networks
• Subject Areas: Highways; Planning and Forecasting; I72: Traffic and Transport Planning;

Filing Info

• Accession Number: 01517445
• Record Type: Publication
• Report/Paper Numbers: 14-2921
• Files: TRIS, TRB, ATRI
• Created Date: Mar 8 2014 10:57AM