The Research Of Speed Prediction Model Of Bus Network Based On Big Data

Nowadays,with the continuous progress of the world economy,it has accelerated the production and transportation level of the transportation industry at home and abroad.The increasing number of motor vehicles and traffic flow in urban areas lead to a lot of traffic safety problems.If the line network speed prediction ability of urban public transport can be improved,the traffic pressure will be alleviated to a certain extent.This is also the research focus of many scholars at present,and there are many pioneering achievements in this field at home and abroad.However,it is not easy to accurately predict the speed of bus network,because:(1)the impact of a combination of factors,the weather alone affects several trends;(2)the amount of data collected by the bus system is quite large and the dimension is relatively high;Therefore,if only the traditional mathematical statistics method is used to process the data,it is difficult to effectively analyze the data of all dimensions,so as to reduce the prediction accuracy.This paper studies the analysis technology based on big data to predict the speed of bus network.The main research includes the following aspects:Firstly,this paper studies the advantages and disadvantages of the methods or algorithms used to predict the speed of bus network at home and abroad.At the same time,it also studies the application of current machine learning algorithms to provide ideas and inspiration for the application of big data technology in this paper.Impact on theory-oriented and speed prediction factors made a comprehensive analysis of the key influence factors of the movement process of the bus,the data statistics,the time data calculation of maximum,minimum,average,standard such as statistics,which will influence the factors are classified,provide theoretical basis for the subsequent data processing;According to the collected GPS information data,IC card swiping data is used for data preprocessing.The relationship between network speed and bus arrival is analyzed,and the prediction of network speed is transformed.Secondly,SVR model and DBN model in machine learning are used for data drilling and prediction.Data sets are selected and refined for data transformation so that they can be applied to regression problems and classification problems respectively.After model training,results obtained from model training are analyzed for comparison in different situations.The experimental results show that the evaluation scales of the two algorithms are enough to realize the training of the model,and the prediction accuracy reaches the expected level.Finally,the advantages and disadvantages of the two algorithms in different situations are analyzed.After comprehensive application,it shows that the two models can forecast the speed of the bus network at a high level,which can alleviate the pressure of urban traffic,and have a pioneering study on the practical problems of transportation on the level of big data.