Od Prediction Of Resident Public Transport Travel Based On Multi-source Data

With the development of science,technology and economy,traffic congestion has become a common problem all over the world.Promoting the development of public transport industry can solve these problems to a great extent.At present,the research on urban public transport planning and optimization has been quite perfect,but it has not been widely used.The main reason is that the initial data obtained is distorted,resulting in that the predicted OD(Origin destination)of residents’ public transport travel can not reflect the real situation of urban residents’ public transport travel,resulting in the loss of practical guiding significance of planning and optimization methods,and the acquisition of accurate OD,which can reflect the real travel situation of urban residents,is imperative.Therefore,the OD prediction method of urban residents’ public transport travel based on multi-source data is developed in this thesis.For the OD investigation and prediction of urban residents’ public transport travel,the principles and characteristics of traditional methods and big data based methods are compared and analyzed;In view of various traffic source data such as mobile phone signaling,bus IC card,GPS and Metro gate data,the methods of data acquisition and processing are discusses respectively,and carries out data fusion,supplement and correction according to the correlation between urban residents’ public transport travel OD matrix generated by various data;On this basis,the wavelet neural network od prediction model is established,and the optimized whale optimization algorithm is used to optimize the model;Take the traffic big data of Furong District of Changsha City as an example to study this model,in which the initial data during the two hours of evening peak in 60 working days are collected.A group of 480 groups of data is set up in 15 minutes,including a training group(360 groups),a testing group(60 groups)and a control group(60 groups).A case study verifies the correctness of the prediction model and the effectiveness of the algorithm.The research results of this thesis have a certain reference value for the application of multi-source data to urban residents’ public transport travel od prediction,and improve the accuracy and timeliness of the prediction.