Short-Term Traffic Flow Prediction Based on Parallel Quasi-Newton Neural Network

Identifying and predicting the situation of traffic flow play an important role in traveler information broadcast and real-time traffic control. In this paper, a short-term traffic flow prediction model based on the parallel self-scaling quasi-Newton (SSPQN) neural network is presented. In this method, a set of parallel search directions are generated at the beginning of each iteration. Each of these directions is selectively chosen from a representative class of quasi-Newton (QN) methods. Inexact line searches are then carried out to estimate the minimum point along each search direction. Experimental and analytical results demonstrate the feasibility of applying SSPQN to traffic flow prediction and prove that it can better satisfy real-time demand of traffic flow prediction.

I Introduction

As an important aspect of Intelligent Transportation Systems (ITS), traffic flow guidance is considered as an optimum way to improve traffic efficiency and mobility. The accuracy of the prediction traffic flow would directly influence the effect of the traffic control and guidance.