| In recent years,China’s high-speed railway has developed rapidly.By the end of 2019,the national railway operating mileage has reached 139,000 kilometers,and the high-speed railway operating mileage is 35,000 kilometers,ranking first in the world.According to the report of national railway construction released by China Railway at the end of 2019,the passenger volume of national railway in 2019 is 3.57 billion,among which the passenger volume of high-speed railway is 2.29 billion,accounting for 64.15%.The high density,high speed,safety,comfort,and punctuality of high-speed railways have made it the mainstream travel of railway passengers.As the basis of passenger flow organization,high-speed railway passenger flow is also the service object of passenger transportation.With the large-scale growth of high-speed railway passenger flow,the expansion of passenger flow data makes the traditional passenger flow analysis difficult to match the passenger flow organization under “demand-driven”,and it cannot meet the diversified and time-varying passenger travel requirements under the high-speed railway network.At the same time,the development and application of big data technology has also brought opportunities for high-speed railway passenger flow analysis in big data environment.The analysis of highspeed railway passenger flow based on big data can more scientifically and comprehensively grasp the law of high-speed railway passenger flow,adjust the train diagram according to the daily travel demand of high-speed railway passenger dynamically,make the train operation plan “one day,one diagram” match accurately the capacity and demand of high-speed railway passenger flow,and improve the service quality and economic benefits of high-speed railway passenger transport.Firstly,this paper summarizes the current situation of high-speed railway passenger flow analysis at home and abroad,and proposes the research of high-speed railway passenger flow based on big data analysis based on the deficiencies of existing research;then this paper elaborates the data preprocessing,data mining and machine learning,and introduces the principles of Apriori algorithm of data mining and Light GBM algorithm of machine learning in detail;then the basic theory is explained for the analysis of high-speed railway passenger flow based on passenger ticket data and passenger occupancy rate data;next the passenger travel behavior,the passenger flow space-time characteristics,ticket allocation status,which are obtained from the analysis of passenger ticket data based on Apriori algorithm,and train operation plan are analyzed one by one,and the optimization suggestions are given for the analysis results.Finally,based on the passenger occupancy rate data,the machine learning Light GBM algorithm of the feature importance analysis of train passenger occupancy rate model is constructed,and then the influencing factors of outbound travel choices of high-speed railway passenger are analysed based on feature importance of the passenger occupancy rate,and obtains the travel choice law of passenger flow. |
