A combinatorial optimization approach to locate traffic counting points in a transport network

We define an integer linear programming model for deriving the complete set of traffic flows on a transport network based on the measurements provided by a suitably located minimal-cost set of counting points. We propose a pair of greedy heuristics that determine lower and upper bounds on the number of traffic sensors. We perform some preliminary numerical simulations in order to evaluate the computational performances of the discussed heuristics. Finally, we briefly discuss the main implications of the proposed model on Origin/Destination (O/D) trip matrix estimation and updating.