Estimating an origin-destination table using a method based on Shortest Augmenting

The most well-known origin-destination estimation methods involve entropy maximization or information minimization. However, these methods often require much more information than is available in reliable form, namely, target trip tables and the extent to which each origin-destination path uses each link. In this paper, a second version of the Shortest Augmenting Path Estimation method (SHAPE-2) is introduced, in which target trip tables and path use proportions are not required. In tests to date, SHAPE-2 has produced good results. It has demonstrated good sensitivity to changes in link loadings on small test networks and, on a real network of special interest to the authors, it performed better than any method previously applied.