Research And Display On Taxi Pick-up Rush Hour And Hotspots Areas Based On Taxi Trajectory Data

With the rapid development of urbanization,the urban population is increasing by years.And the residents' travel needs are becoming more complex,which has caused certain pressure on the traffic.The taxi,as the daily public transport ways for urban residents,can not only meet residents' short-distance travel needs,but relieve traffic pressure and promote the economic development of the city due to its flexibility and convenience.However,the randomness of urban residents' travel and the mobility of taxis lead to the imbalance of information between urban residents and taxi drivers,and taxi supply and demand also imbalanced.To solve this problem,this paper analyzes and excavates the peak times and hot spots of taxi passengers based on Chengdu taxi GPS historical track data,which provides certain reference value for the allocation of urban taxis.The main work is as follows:(1)The preprocessing of trajectory data.In this paper,GPS trajectory data of more than 10,000 taxis in Chengdu were taken as research data.Firstly,the configuration of Spark computing engine is carried out,and then the pre-processing process of trajectory data is realized through Python language.The process mainly includes data cleaning,passenger point extraction,data screening and coordinate transformation,etc.Finally,effective experimental sample data are obtained.(2)The analysis of spatial and temporal distribution characteristics of taxi passenger carrying behavior.Based on the valid sample data of taxi passengers,this paper analyzes the spatiotemporal characteristics of taxi passenger carrying behavior.The passenger load of taxi in different time periods is calculated by mathematical statistical method,and the passenger load peak of working day and non-working day is obtained.Arc GIS tool kernel density analysis is used to display the spatial distribution of taxi pick-up points,so as to obtain the distribution rule of taxi pick-up space.(3)The hot spots analysis of taxi passengers.Based on the characteristic of uneven density distribution of taxi trajectory data,a clustering algorithm about density partition is proposed.By calculating the density peak of trajectory data points,rapid density partition is carried out,and finally the clustering results of each part of the partition are merged and output to obtain accurate taxi passenger hotspot.(4)The system construction based on the taxi passenger peak hours and hot spot area.Based on the above research contents,this paper combines Vue front-end architecture,Docker container and Postgre SQL database to build a taxi passenger load display system in peak hours and hot spots for the users of the system,namely taxi drivers and passengers,which provides a clear and intuitive visual display of the research results.