Abstract
This paper shows an estimation model for day-to-day dynamics of road network flows using partly observed link traffic counts. A least square type model for OD estimation is extended and daily traffic volumes on unobserved links in a network are estimated through a linear mapping of observed link traffic volumes. When some of the links on a network are ceaselessly observed, those data are used for estimating and predicting day-to-day fluctuation of link traffic volumes over the network. The time reliability indexes of the network are formulated using the provided day-to-day fluctuations of link traffic. These models are calculated for observed day-to-day link traffic data in Matsuyama city. The model is available for describing traffic volume on unobserved links and day-to-day fluctuations of link traffic flow over the network.
