Research On Assistant Decision-making Technology Of Train Plan Based On Passenger Flow Big Data

With the increase of railway passenger traffic in China,various problems of transportation scheduling in railway passenger transportation are highlighted.Whether the transportation capacity problem can be solved depends on the train plan.Passenger flow,as the most basic and important factor in railway transport organization,is also the basis for the determining of railway train plan.Therefore,the way to manage and analyze historical passenger flow data more effectively and achieve passenger flow accurate forecast are important to find out the laws of market changes and the trends in passenger demand and the improving of railway's sustainable development capabilities.Big data technology has become a hot topic in recent years.A lot of research and application achievements have been made because of the promotion of big data technology in various trades and professions.Some of the achievements have even changed people's lives.Under the background,the analysis of big data technology and the application of relevant technologies to railway passenger flow data analysis have great research value and exploration significance.This paper mainly studies the technical implementation and realization of big data technology in the passenger flow data analysis and auxiliary decision-making of railway train plan.It aims to improve the passenger flow analysis and forecast technology through big data technology and algorithm ideas.On the basis of the understanding of big data,some possible technical applications for solving railway transportation problems are proposed.Firstly,the paper introduces the basic contents and related influencing factors of the railway train plan,summarizes big data research technologies such as big data visualization,machine learning,and big data processing frameworks,then analyzes the big data technology's application in auxiliary decision-making of the railway train plan.Secondly,the paper analyzes the process and key points of the big data visualization technology in train passenger flow analysis.A passenger flow data sample is visualized by open source visualization framework,the application of interactive visualization in railway passenger flow data is discussed,the temporal and spatial distribution characteristics of passenger flow data are analyzed according to the sample graph.GBRT(Gradient Boosting Regression Tree)in machine learning is widely used because of its strong generalization,while the LSTM(Long Short-Term Memory)neural networks have excellent performance in time series forecasting.In this paper,two types of machine learning algorithms are used to build prediction models.Through the processes of data preprocessing,feature selection,and model construction,the model-based short-term forecast method for passenger flow is presented based on passenger flow characteristics,the prediction effect is verified.The feasibility for the algorithms to achieve accurate short-term traffic forecast models construction is verified by simulation experiments.The application of the model can reduce the workload of passenger flow forecast.Finally,with the application analysis of big data framework,analysis algorithms and relevant technologies,the basic design of train plan assistant decision-making system based on passenger flow big data is proposed,the function and key technologies of each module are explained.