The Research Of Spatial-temporal Causality Based On Traffic Big Data And Its Application In Traffic Flow Prediction

The continuous growth of motor vehicle ownership in China inevitably causes road traffic problems such as traffic congestion.The intelligent transportation system can effectively alleviate these traffic problems,and since its introduction,it has been promoted and developed rapidly in countries around the world.As an important part of the design and application of intelligent transportation system,traffic big data analysis can effectively predict the traffic flow of urban traffic,provide decision support for intelligent transportation system,provide more sensible and coordinated traffic planning for users to travel,improving the efficiency and security of travel by analyzing the relationship between traffic big data and deep causality.In this regard,this paper has carried out the following work:(1)Obtained the traffic dataset and related weather datasets published on the Internet,and carried out data pre-processing such as data cleaning and standardization;on this basis,the time and space characteristics of traffic big data were analyzed.The idea of Granger causality is borrowed and extended,and an analysis method of urban big data traffic data based on spatial-temporal causality analysis is proposed.(2)A traffic flow prediction model based on spatial-temporal causality is established,and a method for quickly screening the influencing factors of traffic flow is proposed.The LSTM network is introduced to construct a spatial-temporal causality analysis algorithm for traffic big data based on recurrent neural network.The experimental analysis proves that the algorithm effectively improves the accuracy of traffic flow prediction.(3)The traffic flow prediction system was designed and analyzed by objectoriented method.The corresponding Android mobile APP and Web server were developed,and the system was tested and tested.Experiments prove that the spatial-temporal causality analysis can effectively extract the key factors affecting traffic flow,combine the key factors extracted with LSTM algorithm to efficiently predict urban traffic flow,and provide decision support for urban management departments and user travel route planning.