Optimal selection and scheduling of road projects by genetic algorithm

The optimisation of road network capacity, the network design problem has previously been studied in small scale networks but little has been done with sequences of investment through a number of years. Genetic algorithms can be used to evaluate this problem dynamically over time. This paper reports on optimal generation of ten-year sequences of road investments for both a rural area and a substantial urban network. The rural evaluation of 34 potential projects modelled linkages throughout Australia but only the local area was specified in detail. There were 68 nodes, 83 links and 36 population centroids. For the urban evaluation of 56 potential projects, there are 782 nodes, 2335 links and 11,928 O-D pairs. For comparison, the urban projects have also been selected and scheduled by ranking according to simple cost-benefit ratio and also by applying rolling or shifting base assessment.


  • English

Media Info

  • Pagination: 18p
  • Monograph Title: Selected proceedings of the 9th World Congress on Transport Research, 22-27 July 2001, Seoul, Korea

Subject/Index Terms

Filing Info

  • Accession Number: 01391111
  • Record Type: Publication
  • Source Agency: ARRB
  • ISBN: 0080442749
  • Files: ATRI
  • Created Date: Aug 23 2012 5:18AM