Prediction Model of Urban Short-time Traffic Flow Based on Intelligent Algorithm

Abstract:
With the development of computer science, intelligent transportation has become the mainstream trend in the field of transportation. The analysis and prediction of urban traffic flow has always been an important basis and research focus in urban traffic planning, traffic safety, traffic facilities, traffic management, traffic control and other aspects. Especially, the analysis and prediction of urban short-time traffic flow is particularly important for urban traffic management and control, and driver travel decision-making, etc., and it is also the basis for the development of intelligent traffic. There are various methods to calculate and predict the short-time traffic flow in smart cities. But with the increasingly towards the complicated traffic system and the pursuit of efficiency of computing, traditional mathematical calculation method is no longer able to meet the needs of urban traffic flow prediction. Intelligent traffic flow calculation method due to its high adaptability and optimization ability training as a prediction model has more advantage gradually, more suitable for the characteristics of modern urban comprehensive transportation system and development. In this paper, the principle and method of back propagation (BP) neural network and new radial basis neural network are studied. The continuous traffic flow situation of urban roads was investigated on the spot. The modeling training and prediction of short-term traffic flow were conducted through two kinds of neural networks based on the actual survey data, and the analysis and evaluation were finally conducted through MATLAB software.