Research And Implementation For Companion Vehicle Discovery Method Based On Spatio-Temporal Trajectory

Intelligent transportation system(ITS)can generate two kinds of typical real-time traffic big data stored by ITS.One of these is GPS data,and the other is the automatic number plate recognition data(ANPR).Research on companion vehicles from the traffic spatiotemporal data has become a hot topic.And it can be applied to traffic management,public security management and other fields.However,considering the real-time,persistent and infinite growth characteristics of real-time trajectory data,it is an urgent problem to find the composition of accompanying vehicles from real-time traffic streaming data in time and accurately.In order to solve the related problems,this thesis studies two kinds of typical real-time traffic big data.A distributed accompanying vehicles real-time discovery method based on incremental calculation is proposed to deal with the license plate recognition data(ANRP).Based on the GPS data collected by on-board equipment,a real-time accompanying vehicles mining method based on time sensitive trajectory similarity measurement is proposed to deal with GPS data collected by on-board equipment.The main contributions of this thesis are as follows:1.This study provides an incremental and distributed approach for discovering traveling companion instantly and continuously based on a data stream of automatic number plate recognition(ANPR).First,a parallelized incremental mining algorithm is designed and implemented in Spark on the basis of traffic-monitoring streaming data.Second,an adjustable data structure DF-tree is proposed that considers the characteristics of companion vehicles with the original ANPR data stream changing dynamically.On the basis of the DF-tree,the system can discover companions without reconstructing the data tree.In addition,we introduce a time decay mechanism to satisfy the spatio-temporal constraints of companion vehicles discovery.Finally,we realize the real-time discovery of companion vehicles based on largescale ANPR data.The proposed methods are evaluated with extensive experiments on real datasets.2.Since that mining GPS real-time data by trajectory similarity measurement algorithm requires high real-time performance,this thesis proposes a time sensitive trajectory similarity measurement algorithm to mine the companion vehicle instantly.First,this thesis introduces the minimum bounding sector(MBS)real-time segmentation algorithm to segment the real-time trajectory.Second,based on the traditional DTW algorithm and LCSS algorithm,the time sensitivity adaptive algorithm is designed.Finally,the distributed flow data processing framework Spark Streaming is used to realize the realtime GPS trajectory similarity measurement algorithm,and the companion vehicle groups are mined in real time.The experimental results show that the incremental mining model and the time sensitive trajectory similarity measurement algorithm established in this thesis are effective for instantly and accurately mining companion vehicle.