A Research On Traffic Volume Forecast Model Based On Neural Network And Induced Traffic Volume

Traffic forecast is utterly necessary for highway construction project,especially during the planning and design stages.Since it is important to determine the construction scale,grade,and traffic networking design,whether the traffic forecast is accurate has a huge impact on the rationality and validity of the decision making and networking design.Only with the accurate forecast,can we effectively avoid the waste of resources engendered by false decision making and better the traffic situation and improve the living standard of people.Based on the influencing factors of traffic,this study compared and analyzed the advantages and disadvantages of the traditional traffic forecasting methods,and pointed out that the drawbacks and limitations of existing traffic volume forecasting models.Moreover,this study analyzed the feasibility of the application of artificial neural network in traffic forecasting and introduced the principals of BP neural network.Based on the existing artificial neural network traffic forecast methods accomplished with MATLAB,this study utilized genetic algorithm to optimize the neural network and applied the induced traffic volume theory to bettering the forecast results and built up a more effective model.The results show that it is rather difficult to predict the traffic volume by using BP neural network alone.The reason is that the BP neural network is easy to fall into the over-fitting problem and over-pursuit of local optimum.However,better accuracy and stability is achieved through utilizing genetic algorithm to optimize the neural network and building up a new forecasting model based upon the GABP algorithm.This new model ought to have a high accuracy on the traffic forecast of roads with stable surroundings of traffic network.However,in terms of roads with expected changes of surroundings,induced traffic correction should be applied for correction.Taking the actual traffic volume of Jinji highway during the 12 th Five-Year Plan period as a case,this paper validated the effectiveness of the model and provided an effective and feasible new method for forecasting the traffic volume of the road reconstruction project in Tianjin.