Trip Table Estimation from Observed Traffic Volumes

This estimation technique identifies the most likely trip table to produce a specified set of traffic counts when assigned to a representative highway network. The resultant table is derived from an entropy maximizing procedure which seeks to define a trip table with the highest degree of disorder or randomness among zone exchanges. Observed traffic counts, network characteristics, and a 'seed' trip table serve to constrain the derivation of this highest disorder table. Trips among all zone pairs are modified in an effort to match observed volumes. Adjustments are made to as many trip table cells as necessary. A package of programs has been written which operate on IBM and IBM compatible microcomputers with a minimum of 512K RAM. Satisfactory results have been achieved in calibrating gravity model derived seed tables and uniform value tables where sufficient count data were available.