Research On Taxi Cruise Strategy Based On GPS Data

As an important part of the urban transportation system,taxis are flexible,fast and convenient to meet the individual needs of passengers and play an important role in public travel.With the continuous development of China’s social economy,the demand for transportation has also increased significantly,and the contradiction between supply and demand in the taxi industry has become increasingly prominent.Due to the uneven distribution of taxi passengers in time and space,the competition in the taxi industry is fierce,and it is difficult for drivers to find passengers.Therefore,how to better recommend passenger-carrying strategies for taxis and guide taxi drivers to efficiently search for passengers is a key scientific problem to be solved to improve operational efficiency and driver income.Based on the taxi GPS trajectory data generated in Lanzhou in April 2021,this paper combines spatial analysis and big data mining methods to explore the value contained in GPS trajectory data,and to provide cruising strategy suggestions for taxi drivers.Firstly,the data preprocessing is carried out on the taxi trajectory data in Lanzhou City,and the taxi trajectory points deviating from the road network are corrected by the geometry-based map matching algorithm to ensure the validity of the data.The extraction algorithm is designed to obtain the OD information and related operation indicators of taxi pick-up and drop-off,and to mine the travel behavior information of taxis.On this basis,the spatial and temporal characteristics of taxi passenger distribution in Lanzhou are analyzed.Among them,the time distribution of travel demand is mainly explored from the characteristics of the total amount of taxi trips,the distribution of demand in each time period,the length of passengers,the waiting time and the income of passengers;in terms of the spatial distribution characteristics,the analysis of working days and non-working days is carried out.The hot spots of taxi pick-up and drop-off points at different peaks are used to study the passenger flow migration in major urban areas.Secondly,study the operation mode of taxi drivers in a complex metropolitan environment,consider the relevant factors affecting the income of taxi drivers,establish rules to divide taxi drivers into three types of drivers with high,medium and low income,and compare different types of drivers.Using a multi-class ordinal Logit regression model to study the influence of factors such as no-load passenger distance,passenger distance,passenger speed and other factors on driver income,and to provide suggestions to cruising taxi drivers from different perspectives;using density-based The DBSCAN algorithm mines the passenger hotspot areas of high-income drivers in various peak hours,provides middle-income and low-income drivers with a reference on the passenger space,and reduces the income gap.Finally,Considering factors such as passenger carrying probability,passenger source benefit,passenger-seeking path cost,and road traffic conditions in hotspot areas,a taxi-seeking recommendation model is constructed.The actual data of Lanzhou City(the morning peak on April 9,2021)is used to guide empty taxis to search for customers.The results show that the model established in this paper has a certain guiding role for taxis to search for customers,and can improve the income and operation of taxi drivers.benefits,etc.are of great significance.