Optimal Use of Plate-Scanning Resources for Route Flow Estimation in Traffic Networks

This paper discusses the problem of optimizing the use of scanning resources for route flow estimation in traffic networks. Three problems are considered: 1) Minimize the number of cameras to be used for estimating a given subset of route flows; 2) identify the subsets of links to be scanned for a given number of available cameras, assuming that several runs can be done; and 3) solve the previous problems with consideration of scanning errors and error recovery. In addition to the main objective, which consists of minimizing the number of cameras, minimizing the total number of links to be scanned in all runs and maximizing the number of determined route flows are also considered as secondary and tertiary objectives, respectively. A simple and a medium-size real network are used to illustrate the proposed methods and prove the feasibility of the proposed methodology.