An evaluation of statistical performances of O/D matrix estimators combining results of investigations and/or models with traffic counts

This paper presents an analysis of the most known methodologies for estimating and/or updating origin-to-destination trip matrices from traffic counts. Effects due to variations of some significative parameters on statistical performances of these estimators are compared by adopting simulations on a test network. An evaluation of the difference, in terms of cost, between two different estimating procedures is also carried out.