A Bi-Level Programming Model for Origin-Destination Estimation Based on FCD

The aim of this paper was to study the floating car data (FCD) application from origin to destination (OD) estimation issues. Thus the study put forward a method to estimate the OD matrix based on FCD. In this study, we first propose a model for obtaining prior OD matrix from FCD. Then a bi-level programming model was designed for OD flow estimation. As mentioned above the bi- level programming model is multi-objective optimization model, and the second model is for obtaining prior OD matrix which was integrated into bi-level model. The first model is a time-dependent OD traffic assignment model. Two-stages of genetic algorithm were used for solving the bi-level programming model. At last, we apply this model to a test network and examine its estimation performance. Compared with the traditional estimation method without FCD information, the OD estimation result was found to be more accurate and efficient.