Analysis And Research Of Highway Traffic Flow Prediction Model

In recent years,Intelligent Transportation System(ITS) has been made more and more attention.It was given birth to the background of contemporary science and technology fully developed which seeks to introduce advanced computer technology, communication technology,database technology and Artificial Intelligence transportation so as to solve the traffic congestion,ensure the safety and improve the using rate of traffic network.ITS involves many aspects of traffic field.And in these aspects,one important application is traffic prediction in highway transport.The main concept of traffic prediction is to provide the best route,balance traffic flow,optimize traffic management and improve traffic control by the use of appropriate methods of dynamic prediction and the information of cross roads and intersection.The paper discusses and researches several methods of traffic flow prediction in the history at home and abroad considering the feature of traffic flow. And it lists advantages,shortcomings and the conditions of application by comparing these methods.Then,it applies isolation niche Genetic Algorithm to optimize the neural network in order to construct reasonable neural network prediction model aiming at the characteristic of BP neural network operating and by considering current specific traffic flow,history traffic data and upstream traffic flow data.Finally,Use this model to forecast the traffic flow of highway section of urban road,have made the comparatively satisfactory result.