Dynamic od matrix estimation method based on monitoring data

The invention provides a dynamic OD matrix estimation method based on monitoring data, which belongs to the field of intelligent transportation. A traditional method does not take into account the sparsity of some actual road network OD matrices, and the calculated OD matrix and the traffic distributed on a road network are deviated from the actual. In a solution process, distributing the OD matrix to the road network takes a certain time. Especially when a road segment is in congestion, a traditional OD matrix distribution algorithm takes a long time, which is difficult to meet real-time requirements. According to the invention, a dynamic OD matrix estimation model under sparse constraints is constructed, and a fast dynamic OD matrix estimation algorithm under sparse constraints is proposed; an OD matrix distribution function is fitted based on a radial basis neural network; on the basis, key problems to be solved comprise the dynamic OD matrix estimation model under sparse constraints and the real-time dynamic OD matrix estimation algorithm under sparse constraints; and in theory, the proposed scheme can accurately estimate a dynamic OD matrix to meet real-time requirements.