Research On The Identification Method Of Suspected Illegal Passenger Vehicle Based On The Highway Toll Data

Public transportation system,mode and passenger transportation management are still not perfect in China.It's unable to meet the diverse needs of residents' travel and creating market opportunities for illegal passenger vehicles.At present,the illegal passenger transport has not been effectively controlled,which has seriously affected the management of the passenger transport market,disrupted the public order,endangered the safety of passengers,and damaged public rights.The main reason is that the traditional method of identifying illegal passenger vehicles requires a lot of manpower,material resources and time.And it is inefficient.Therefore,it is an urgent problem to study a scientific method that can quickly and effectively identify illegal passenger vehicles in a large area,thereby improving law enforcement efficiency and effectively suppressing or eliminating illegal passenger transport.And the formation of big data theory and intelligent transportation has brought great opportunities to solve such problems.The investigation and handling of illegal passenger vehicles can be divided into two aspects: the discovery of suspected illegal passenger vehicles and the punishment of evidence.Due to the high concealment and wide distribution of illegal passenger vehicles,it is a key and difficult point to quickly discover suspected illegal passenger vehicles on a large scale.Through theoretical analysis and in-depth investigation,this paper studies the status of illegal passenger transport and reveals the characteristics of illegal passenger vehicles in terms of travel space and time,including the five characteristics of “travel intensity”,“specific line travel ratio”,“specific line travel intensity”,“weekend travel intensity” and “morning and evening peak travel ratio”.Those are used as indicators to identify suspected illegal passenger vehicles.Five sets of eigenvalues for identifying suspected illegal passenger vehicles are established based on highway toll data.Under this,two methods for identifying suspected illegal passenger vehicles based on spatial features,spatial and temporal characteristics are established combined with the theory of clustering algorithm.The DB clustering evaluation index is used to analyze the clustering effect of K-MEANS algorithm and DBSCAN algorithm as the basis for selecting clustering algorithm.The case analysis and verification on the two methods for identifying suspected illegal passenger vehicles are analyzed using the information of highway toll data of a city.The verification results show that the accuracy of the illegal passenger vehicles identified by the two methods is all relatively high,at least eighty percent.The identification methods of suspected illegal passenger vehicles is effective and has good practicability.Finally,the corresponding suspected illegal passenger vehicle identification system is developed.In summary,the results of this paper can provide a scientific basis for the government transportation management department to control illegal passenger transport,and provide effective methods for traffic law enforcement departments,which can significantly improve the efficiency and effectiveness of law enforcement.