A Metaheuristic Dynamic Traffic Assignment Model for O-D Matrix Estimation using Aggregate Data

For a static/dynamic O-D matrix estimation, usually, the basic required information is a starting estimation of O-D matrix and a set of traffic counts. In the era of the Intelligent Transportation Systems, a dynamic estimation of traffic demand has become a crucial issue. Different Dynamic Traffic Assignment (DTA) models have been proposed, used also for O-D matrices estimation. This paper presents a dynamic O-D demand estimator, using a novel simulation-based DTA algorithm. The core of the proposed algorithm is a mesoscopic dynamic network loading model used in conjunction with a Bee Colony Optimization (BCO). The BCO is capable to solve high level combinatorial problems with fast convergence performances, allowing to overcome classical demand-flow relationships drawbacks.