A method for estimating vehicle od in road network based on sampled trajectory data

Abstract

The invention relates to a road network vehicle OD estimation method based on sampling trajectory data, which comprises the following steps: 1) sample vehicle trajectory data are preprocessed as trajectory vectors, a prior road section traffic flow matrix and a road section travel time matrix are constructed, and a prior OD matrix is calculated and obtained; 2) the path selection proportions of the sample vehicles are aggregated as a whole road network distribution matrix, and through estimation, a BPR function of each road section calibrated by the road section traffic flow and the road section travel time is obtained; and 3) a generalized least square framework is extended, the OD, the road section traffic flow and the road section travel time are together taken as decision variables, atraffic assignment relationship and the BPR function are taken as constraints, and an optimization model is solved by a gradient descent method. Compared with the prior art, the method disclosed in the invention has the advantages of single-source data, reliable estimation results, dynamic estimation and good applicability and the like.

Description

technical field [0001] The invention relates to the field of road network traffic state estimation in the field of intelligent transportation, in particular to a road network vehicle OD estimation method based on sampling trajectory data. Background technique [0002] Urban road network vehicle OD, that is, the aggregation of path flows with the same origin and destination points in the urban road network, is one of the most important parameters of traffic status at the road network level, which can directly reflect the traffic demand of the road network in the current period. On this basis, research on travel time estimation, congestion prediction, and route planning of urban road networks can be carried out. [0003] In the past few decades, a large number of studies have used fixed-point detection data (coils, geomagnetism, AVI data, etc.) to estimate the OD of the urban road network, and reversed the OD of the road network through the flow detection data of some road sec...

AI Technical Summary

Problems solved by technology

[0004] On the other hand, most of the existing OD estimation analysis models rely on the prior information of the road network, such as: road network OD information obtained through traffic surveys or historical flow information about some paths in the road network. On the one hand, these data are limited to Part of the road network, on the other hand, its timeliness is also poor; and estimation methods based on machine learning often require a large amount of multi-source data for training, and often the model generalization ability is poor and does not have transferability

Images