The Research On Short-term Traffic Flow Forecasting

As a hot research topic in the field of intelligent transportation, short-term traffic flow forecasting is one of the key technologies for intelligent traffic control and traffic guidance.Short-term traffic flow has characteristics, such as nonlinear, time-varying, uncertainty, instability. Respectively, this paper research the short-term traffic flow forecasting in the view of linear system theory and nonlinear systems theory. Through analysing the autocorrelation and partial autocorrelation function of short-term traffic flow time series, a specific time series model is established, and a prediction is conducted using the regression model based on ARMA; At the same time, considering the short-term chaos effect of traffic flow firstly, this paper computes the embedding dimension and delay time of phase-space by C-C algorithm, on this basis, reconstructs the phase space, and then uses small amount of data algorithm to obtain the maximum lyapunov exponent, so the chaotic characteristic of the short-term traffic flow can be analized, finally, in the premise of chaotic characteristics, uses the forecasting model based on the largest lyapunov exponent to conduct the forecasting of short-term traffic flow.Finally, this paper uses MATLAB programming to realize the modeling and simulation of these two prediction models, and does the error analysis between the the predictive value of model output and the true value, to obtain good results. Shorten the sampling period with the traffic flow, the chaotic effect is even more obvious, and the advantages of the forecasting model based on the largest lyapunov exponent will become more pronounced.