The short-term traffic flow prediction based on neural network

As we all know, to predict the short-term traffic flow accurately and efficiently is the premise and key of traffic management and control. Based on these existing study, this paper selected BP neural network model in which the traffic flow difference was taken as the input parameter, applied the thought of dynamic rolling prediction to design a new short-term traffic flow prediction method, and wrote the corresponding program. Then using the actual observation data of traffic flow presented the model structure, thought and calculation steps of this new method. The results show this method is feasibility, reliability, and of some practical value.

I. Introduction

As urbanization progresses, the traffic demand increasingly expands, while the urban land resources are limited, which led to traffic jams daily occurrence, traffic accidents high frequency. To resolve these traffic problems scientifically and reasonably has become a society-wide consensus, while given the domestic urban governance in recent years, only building transportation infrastructures to cannot relieve the traffic pressure. One important way to increase transport efficiency, reduce traffic congestion and improve traffic safety situation, is to implement traffic guidance and control, effectively use the road resource and give full play to vehicle function. Furthermore, the traffic flow prediction is just the premise and key of traffic management and control.