Multi-path Recognition System Based On Big Data In Expressway Domain

With the rapid development of China's expressway,the road network has become complex with many emerging loops.In the road network,vehicles can be driven between the entry and exit toll stations.The multi-path problem makes it difficult to determine the actual travel path of the vehicle.Finding the actual driving path of the vehicle in the expressway network multiple paths accuratly not only benefits the reasonable charge and expense allocation,but also the expressway users and managerial departments.The identification stations equipped with license plate camera systems or RFID equipment in the expressway network could monitor vehicles and upload data in real time.Combine with high-speed toll station data can identify vehicle driving path.However,these devices are affected by the lack of ambient brightness,transparency,and hardware device failures may result in the lack of vehicle identified data,which makes it difficult to identify the path effectively.How to use the real-time charge data and identified data to identify the multipath of the vehicle has become an important issue.In this paper,a multi-path recognition system based on expressway big data is researched and implemented.The main work of the thesis includes:1.A accurate identification method of real-time online for multipath based on spatio-temporal relationship is proposed.Aiming at the accuracy and real-time requirements of multi-path recognition,uses vehicle charging data and identified data to design a real-time online accurate identification method based on spatio-temporal relationship.The real-time charging data and identified data are processed accordingly.The temporal and spatial correlation of the vehicles in the road network determines the travel path of the vehicle.2.A probability identification method for expressway multipath based on historical data is proposed.The identified data on the expressway is missing due to environmental disadvantages and equipment failures,which makes it difficult to accurately identify the multipath.For this problem,designs a probability identification method for expressway multipath based on historical data,which is based on the segment-based aggregation performs statistical calculation on the historical traffic data of the vehicle,obtains the traffic probability of each road segment,then combines the greedy algorithm to give the ambiguous path recognition process under the condition of data loss.3.A multi-path recognition system is designed and implemented.Based on the above-mentioned path recognition methods.Experiments based on real data show that the path recognition methods in this paper have better performance and higher accuracy than the shortest path identification method.