Statistical Analysis On The Passenger Capacity Of Taxi In A City

The urban taxi passenger transportation system is a means of transportation for people’s daily travel,and it is also a core component of the urban transportation system.With the continuous growth of urban passenger traffic in my country,the number of taxis has shown a slight upward trend,but the overall passenger volume has fluctuated and declined,and there has been a mismatch between the number of passengers and the number of taxis.Therefore,it is of great significance to study the law of taxi capacity.This thesis intends to conduct a statistical analysis of the operating status of taxis in Liuzhou City,and summarize the regular characteristics of passenger flow in time and space from two different modes during the day and night,so as to provide a more effective operating strategy for the taxi market.Firstly,based on the taxi GPS positioning data in Liuzhou from June 2019 to November 2019,the original data set is preprocessed to extract the locations of the pickup and drop-off points,and then the calculation formula for the passenger capacity is given.The data is converted,and a descriptive statistical analysis is made on the passenger carrying situation of the taxi from the three aspects of passenger carrying month,passenger carrying time period and passenger carrying area.Secondly,it discusses the influencing factors of taxis carrying passengers from two aspects: the strategy of looking for passengers and the strategy of sending off passengers.According to the established index system,the research of multiple orderly logistic regression is carried out.Finally,using k-means clustering method and kernel density analysis to deeply explore the time and space distribution characteristics of the two groups of passengers during the day and night,and comparing the hot spots of the two groups of data at different time periods to find out their corresponding influencing factors.This dissertation constructs an index system of factors affecting taxi passenger carrying.Based on a multivariate ordered logistic regression model,it quantifies the impact of five indicators on taxi passenger carrying distance,boarding intensity,patrol/waiting passengers,roundaboutness,and drop-off speed.Based on the clustering of the two sets of data during the day and night,four characteristic time periods with significant performance were obtained,and the passenger hotspots in each characteristic time period were extracted,and the experience was explored.The differentiated business model caused by different influencing factors has improved the passenger carrying efficiency of taxi drivers and provided big data support and services for the layout of traffic flow.