Spatial And Temporal Characteristics Analysis And Regional Distribution Prediction Of Illegal Taxis Using License Plate Recognition Data

As an emerging technology,the check point video license plate recognition facilitates obtain a large amount of information on vehicle operations,and it plays an important role in the analysis of vehicle operational laws.The illegal behaviors of taxis have a serious negative impact on social security and normal business order,and also put forward higher requirements for industry supervision and inspection.This thesis focuses on analyzing the snapshot data at the check points and historical track data of illegal taxis to have a better understanding of their general temporal and spatial distribution characteristics and assess and visualize the hotspot areas haunted by illegal taxis.In addition,a prediction model of the regional distribution of illegal taxis is constructed to predict the regional distribution of illegal taxis in the future.It is conducive to the directional arrangement of law enforcement forces by the traffic management department,the implementation of targeted checks,and the formation of an illegal operation governance pattern of accurate,efficient and targeted governance.The detailed researches of this thesis are as follows.（1）According to the video information obtained by the intelligent monitoring equipments of the check points in Guangzhou,a structured database including license plate number,check point number,check point name,coordinates,and passing time are established.The author analyzes various abnormalities existing in the obtained license plate recognition data and filters the original data,so as to obtain a high-quality data set and ensure that the data can truly reflect the operation status of illegal taxis.（2）Throught the preliminary data analysis,a fundamental understanding of the situation and overall posture of illegal taxis is obtained according to travel frequency,time distribution,and regional distribution.According to the frequency and number of illegal taxis captured by different check points,the spatial distribution of illegal taxis is analyzed.In addition,the overall distribution law of illegal taxis in different time periods and the comparative distribution law of different time periods on the same date are analyzed according to the records of passing times at check points.（3）Based on the above study of the temporal and spatial distribution characteristics of illegal taxis,DBSCAN clustering algorithm is adopted to conduct clustering analysis on the trajectory point data of illegal taxis.Then the kernel density analysis method is used for hotspot assessment and the clustering results are visualized by Arc GIS software,so as to identify the hotspot areas of illegal taxi travel within the research scope.（4）On the basis of studying the temporal and spatial distribution characteristics and hotspot areas of illegal taxis,comprehensively considering the historical trajectory data,a regional distribution prediction model of illegal taxis is established by using Markov process prediction algorithm.The prediction effect of the model is verified through simulation prediction experiments.These are of great significance to improve the accuracy of inspection and control of law enforcement departments.