Data-driven Analysis Of Bus Travel Characteristics And Operation Efficiency

Public transportation is an important basis of urban economic and social development services.Public transport is major to the urban development,directly related to the urban economy and the residents' life.There is the guide line on urban economy and significant role in the overall importance.What's more,it can effectively alleviate the urban traffic congestion problems,significantly improve urban people's living environment,reduce vehicle emissions.In view of the unreasonable allocation of public transport resources,the unscientific planning of public transport routes and the low efficiency of public transport problems related to public resources should be scientifically evaluated to ensure the convenience of travelers,with the same time,it can improve the economic and social development of the city.In the modern information age,the application of various traffic detection devices produces a large amount of traffic data,so it is reasonable to develop and utilize relevant traffic data resources.Traditional traffic operation optimization relies on manual experience to make relevant adjustments,but it is often unscientific and unreasonable,and it may not be able to effectively and quickly analyze the existing problems.In this paper,first of all,from the current circumstances of road public transit,on the premise of data-driven,transportation IC card data and GPS data as the foundation,combined with machine learning algorithms.Data can get time and travel point.Though screening effective public transit data,then specifies the transit OD traffic routes.Second,the analysis of different sites of different periods in bus lines travel traffic,passenger flow prediction could carry out on the site and end point in the data.Drawing a bus travelers laws of travel time and travel site gathered,through a fixed place of travel time and travel site surrounding land nature.Above all we can combine people's travel intentions related analysis,draw commuters travel rule changes,including travel peak,travel with the working days difference in working days.Travel characters,it is concluded that the combination of bus lines of a part of the site traffic analysis.The purpose can do bus stations of scientific prediction,based on machine learning,neural network algorithm and the bus stops near the nature of the land to build predictive model algorithm.The model can make more accurately calculating the corresponding time period under different time of work site traffic.Based on the above analysis of bus travel characteristics and the prediction of passenger flow at the station,the following evaluation of bus operation efficiency is conducted.So the utilization of public transport resources and the travel time of bus passengers are reasonably analyzed.The distribution law of urban areas can be obtained through above analysis.The evaluation system of bus benefit will deeply study.It can combine with bus punctuality in the bus system and vehicle based on GPS data to evaluate emissions.With Qingdao city bus investigation,analysis and bus flow,we can make a reasonable scientific evaluation method for urban public transport efficiency.The paper's bus benefit evaluation system research can provide reference for related research.