Akademik Veri Yönetim Sistemi
PROCEEDINGS OF THE INSTITUTION OF CIVIL ENGINEERS-TRANSPORT, cilt.164, sa.1, ss.33-42, 2011 (SCI-Expanded)
The usability of the Taguchi method in traffic simulation was investigated to demonstrate whether full factorial scenario testing is necessary on a small road network by changing certain links into one-way streets to ascertain the minimum total delay. A network with 52 links was constructed and four arbitrarily chosen links were made one-way streets, one by one in both directions. The best directions of traffic for each of these four links which gave minimum total delay time over the whole network were studied by means of Paramics microsimulation software. The simulation runs were designed using an orthogonal array technique in L(8) array with four factors (with two levels each) and three interactions. The total delay time was selected as the response variable. The response data were analysed using an analysis of variance technique by the Taguchi method. In addition, the Taguchi outcome was judged against the results of the full factorial design of the actual Paramics runs for the best and transparent comparison. Furthermore, in comparison with full factorial design, the results demonstrated that the use of the Taguchi method enabled the interactions between the factors to be seen, which cannot be achieved so simply by simulation trials.