Methods for estimating trip tables from traffic counts are potentially useful because of their relative efficiency in data requirements. Two techniques for estimating existing trip tables in urban highway networks--the information theory (IT) technique and the LINKOD model--are analyzed in this paper. The separate description of the two techniques is followed by a formulation of an algorithm that is designed for application of the two techniques as well as other variations. By using the allgorithm, extensive experimentation with the various techniques is made by using artificial data. Both the convergence speeds and the ability of the techniques to stay close to the target trip table are evaluated. The main contribution of the paper is its presentation of the two major techniques within an easily understood, unified format. It opens a way for extending the IT techniques for equilibrium assignment problems. (A)
Abstract