Spatial Distribution Pattern And Analysis Of Macroscopic Influencing Factors For Taxi Accidents

As one of the common modes of travel for urban residents,taxicabs have always been popular among travelers due to their convenience,speed and comfort,forming an urban passenger transportation system together with other modes of transportation.However,with the increase in the number of traditional taxicabs in cities and the emergence of online vehicles,the rate of taxi accidents has gradually increased,posing a great challenge to traffic safety in urban areas.Although the traditional traffic safety analysis methods are simple and easy to understand,they lack macro-level control of the traffic system and are essentially "post-event remediation".In recent years,scholars have started to introduce the idea of "traffic safety planning",the macro-level planning elements have been incorporated into the scope of traffic safety research.Using the traffic analysis zone(TAZ)as the basic unit,analyzing the key factors affecting traffic accidents at the macro level has become a research hotspot.Based on the above background,this paper firstly uses nuclear density analysis and spatial autocorrelation analysis to study the spatial distribution characteristics of cab accidents.Secondly,based on the macro traffic safety research framework,the traffic analysis zone(TAZ)is taken as the unit,and the macro factors such as cell population,economy,land use properties,and road characteristics are considered comprehensively,and a Bayesian conditional autoregressive model is constructed to study the influence of the above factors on the cell cab.Finally,relevant suggestions and measures to improve and enhance cab traffic safety are proposed based on the analysis of the model results.The main contents are as follows.(1)Analysis of the spatial distribution pattern and characteristics of taxi accidents.Firstly,the global spatial autocorrelation analysis method is used to verify the existence of obvious spatial autocorrelation in the spatial distribution of the neighborhood taxi accident rate;secondly,the local spatial autocorrelation method is used to analyze whether there is spatial heterogeneity in the neighborhood cab traffic accident rate within the study area;Finally,the kernel density analysis method is used to visually analyze the spatial density distribution of the taxi accident rate in the study area,identify the high-density accident areas,and initially analyze the reasons for the higher accident rate.(2)Analysis and modeling of macroscopic influencing factors of taxi accidents.Since the taxi accident rate in the district has spatial autocorrelation,the Bayesian residuals are introduced on the basis of the negative binomial model,and the spatial weight matrix is calculated based on the spatial geographic distance and cab travel connection respectively,so as to construct a Bayesian conditional autoregressive model and analyze the key macro factors affecting the taxi accident rate in the district based on the model results.(3)On the basis of the above analysis,measures and suggestions for improving urban taxi safety are proposed from three levels: planning department,traffic control department and taxi company management.This research was funded by the National Social Science Foundation of China(19BGL239).