Peer vehicle discovery method based on streaming large data

Abstract

The invention claims a peer vehicle discovery method based on streaming large data. The method comprises the steps: first, clustering the bayonet traffic and the number of branches. Through the obtained bayonet clusters, the role identification of the bayonet is verified, and the power law of the bayonet traffic is verified. Secondly, the Spark streaming time sliding window is introduced on the basis of streaming data. According to the driving trajectory, the context between the vehicles is obtained, and the creation and improvement of the peer corpus is completed. Finally, the PDGC (Plate number Dynamic Graph Computing) algorithm is proposed. A dynamic relationship diagram between vehicles is established based on dynamic corpus and bayonet role identification. The bayonet role is associated with the map between the influence factor and the vehicle, and the peer train is obtained by real-time trimming of the vehicle map and calculation of the weights between the peer vehicles, effectively reducing the complexity of data processing. The method is provided with the ability to discover peer vehicle groups in real time, which is not only used for searching for similar tracks, but alsois possible to mine the tracking vehicle by calculating the out-degree and penetration of the vertices of the vehicle map.

Description

technical field [0001] The invention belongs to the field of big data mining, mainly relates to the field of intelligent transportation, in particular to a method for discovering peer vehicles based on big data. Background technique [0002] With the advancement of mobile devices and recognition technology, a large amount of trajectory data has been recorded, and these data are intensively used for trajectory clustering, traffic management, outlier detection, area of ​​interest, privacy protection, location recommendation, etc. There are two types of trajectory data sources, one is from external equipment: the moving object information data captured by the bayonet probe, which records the characteristics of the moving object. Another type of trajectory data is generated by the moving object itself: the positioning data generated by the mobile device on the pedestrian, the GPS data generated by the vehicle's own device, including the location information of the moving object ...

AI Technical Summary

Problems solved by technology

Efficient structured programming models have developed many graph computing frameworks, and they usually follow the coded graph to handle vertex parallelism and communication between edges. Computation and storage of data streams consume a lot of resources.

Images