Expressway Short-term Traffic Flow Forecast

With the continuous development and construction of the Expressway,it has been chosen by more and more people as the main way of daily travel because of the fast and effective traffic capacity of it.However,the existing capacity of traffic network is far from enough to meet the needs of increasing travel demand of people,then great traffic congestion has occurred of which the consequence is the frequent occurrence of traffic accidents and the deterioration of environmental pollution.Therefore,the study on the theory and method of Short-term traffic flow forecasting is an essential prerequisite to achieve reasonable and effective traffic guidance,alleviate traffic congestion,reduce traffic accidents and improve environmental pollution of expressway.Acquisition technology and equipment of expressway data continue to improve,it is possible to forecast the short-term traffic flow of expressway.In this paper,different prediction models are established,and improved,and the prediction results are compared base on the predictability of short term traffic volume on expressway.First of all,this paper summarizes the research background,significance and research status of short time traffic flow forecast methods,deficiencies in various forecasting methods and models are analyzed,and effective collection methods is applied to count the short-term traffic flow of Lan Hai and Wu Guan two expressways of Gansu province for practical analysis of subsequent models.Secondly,some basic concepts and parameters of Chaos theory are introduced based on the internal characteristics of short-time traffic flow time series.In order to analyze the time series,the phase space is reconstructed.In order to dig out the actual rules inside of its,the C-C algorithm is used to calculate the delay time and embedding dimension of the two sets of experimental data of Lan Hai and Wu Guan expressway which is used to reconstruct the phase space of the original time series.On this basis,the maximum Lyapunov exponent of the two sets of data is calculated by the method of small data,and the calculated results are positive which indicates Chaos theory can be used for corresponding analysis and research of these data.Then,the concept of artificial neural nets is introduced and the wavelet neural network and the RBF neural network were used to predict the short-term traffic volume.Prior to this,the number of neurons in input layer and output layer of neural network is designed reasonably by using the delay time and embedding dimension of the reconstructed two sets of data respectively,then a good network topology is established to predict the collected data of Lan Hai and Wu Guan expressway in the created network.The analysis and calculation of experimental result show that the prediction effect of RBF neural network is better than that of wavelet neural network.Finally,aiming at the shortcomings of these two neural networks,genetic algorithm is used to optimize the initial parameters of them in order to ensure that the output of the network is better.After the prediction experiment of two groups of data collected from Lan Hai and Wu Guan expressway,It can be get that the prediction error of two modified neural networks have been improved,simultaneously,the improved RBF neural network prediction model is better than the improved wavelet neural network prediction model which can predict the short-term traffic volume better.