Research On OD Analysis Based On Positioning Data Of Electric Bicycle

The widespread application of intelligent transportation systems can make full use of road conditions,reduce traffic congestion greatly,increase traffic volume,and fundamentally solve traffic problems.An important basic data of the intelligent transportation system is the transportation trip OD(Origin Destination)matrix.The transportation trip OD matrix contains the travel information of the residents,and the travel demands of the residents can be obtained.The current urban travel OD matrix estimation data and methods have two limitations:(1)The OD estimation method can only handle dense trajectories with relatively small and fixed time intervals between trajectory data points,and does not support data with relatively large and variable interval between trajectory data points;(2)The OD matrix obtained from the travel positioning data of a car or bus is often the travel volume from a large range of partitions to partitions.However,electric bicycles have a short travel distance due to their power limit,and it is necessary to select a specific location as the starting point of travel to estimate the OD matrix of short-distance electric bicycle travel.Therefore,this paper studies the OD matrix estimation technology based on the positioning data of electric bicycles with irregular time intervals.This article first explains the research background and significance of OD estimation,and introduces the current research status of OD estimation from two aspects: static estimation and dynamic estimation.Then,it introduces the related theories of OD estimation,including the definition of traffic travel OD matrix and common algorithms for dividing trajectory stop points.It points out that each algorithm is affected by algorithm threshold selection and trajectory data characteristics,resulting in inaccurate divisions.Then,starting from the data characteristics of electric bicycles,this article proposes a time-constrained space based on TDBSCAN(Timetable-Density-Based Spatial Clustering of Applications with Noise)for the problem of dense or sparse positioning caused by the unfixed time interval of electric bicycle positioning data.The clustering algorithm divides the trajectory stop points,and filters and optimizes the clustering results at the time level to obtain the user travel stay points and the travel chain,and obtain the user’s electric bicycle travel OD matrix through statistics.This paper selects the travel trajectory of a single user and multiple users to divide the stop points,and compares the division results of the algorithm in this paper and the DJ-Cluster(Join-Based Clustering)algorithm by using the confusion matrix and trend changes to verify the effectiveness of the algorithm in this paper.Finally,this paper selects the all-day electric bicycle travel trajectory of users in a certain city,divides their stay points,obtains the travel OD matrix,and analyzes the traffic flow of electric bicycle travel in this city.