Research of Short-terra Traffic Flow Forecast Method Based on Nonparametric Regression

摘要

In view of the strict requirement of short-term forecast accuracy and real-time of traffic control and traffic guidance, and take the nonlinear and uncertain characteristics of traffic flow into consideration. Then based on nonparametric regression, this paper improved the nonparametric regression forecast method with the combined correlation coefficient and trend method. Not only the calculating amounts in prediction process reduce, but also can ensure that it has higher prediction accuracy. The Improved model provides some new ideas and methods for short-term traffic forecast.