Short-term Forecasting Model For Urban Traffic Flow

With the improvement of economic level, the burden of the family reducing and the purchase of private cars increasing, traffic jams are plagued by a big traffic problem of China and even in today’s international, how to alleviate the pressure of the traffic become a problem to be solved in our country. To obtain accurate real-time traffic flow is the basis of induction and traffic control, but also the key to solve the problem of all kinds of traffic.Based on short-term traffic flow prediction and the analysis of research status at home and abroad, analyses the characteristics of urban traffic flow, summarizes the existing prediction method, is proposed based on phase space reconstruction of kalman filter traffic flow forecast simulation model. To get hidden in the property of the short-time traffic flow one dimensional time series, time series of one dimensional reconstruction, C-C algorithm is used to determine space of delay time and embedding dimension. This model is based on phase points of the phase space reconstruction as state vector phase point description, using the Kalman filtering theory in real time prediction and correction phase point of future evolution. Based on the two theories of short-term traffic flow forecasting model, at last, according to the actual traffic conditions in a certain section of simulation.Theory of support vector machine SVM careful research and analysis, and to determine the type of kernel function for predicting object of this paper, for the lack of training methods introduced before the data wavelet denoising theory, according to the characteristics of several wavelets, the choice to join the compromise factor the thresholding method for data de-noising, and in order to improve the prediction accuracy by ant colony optimization algorithm to optimize the parameters of the model, constructed prediction model parameter optimization of SVM, and simulation analysis the actual traffic flow, validation based on the availability and practicability of the algorithm.To compare analysis, will build the phase space reconstruction of kalman filter traffic flow forecast simulation model and the SVM based on parameter optimization of short-term traffic flow prediction model for simulation, and by using the proposed index information comprehensive comparison of two groups, the simulation results show that the parameter optimization of the SVM model based on intelligent algorithm is more effective to improve the prediction precision of the traffic flow, and prove that the smart combination algorithm can achieve better prediction effect.