Short term traffic flow prediction method for urban roads: improved Bayesian network

Short term traffic flow prediction on urban roads is an important issue in traffic planning and management. In order to improve the accuracy and efficiency of short-term traffic flow prediction, the paper proposes to apply an improved Bayesian network to the research of short-term traffic flow prediction methods for urban roads. Firstly, perform singular spectrum analysis on short-term traffic flow data of urban roads to reduce noise in the data. Secondly, construct the input data matrix of the convolutional neural network to extract short-term traffic flow features and reduce feature size. Finally, the short-term traffic flow prediction of the dynamic Bayesian network model is completed by calculating the predicted value corresponding to the maximum posterior distribution probability based on the extracted short-term traffic flow characteristics. The experimental results show that the short-term traffic flow prediction accuracy of this method is relatively high.