Research on improved algorithm for traffic information prediction

Predictions for traffic conditions, especially short-term prediction, is one of the central topics in traffic information simulation and controlling relevant research and practice. However, it is not easy to select the most appropriate traffic prediction method for one particular application. In this paper we select two typical methods in this field, then compared about the advantages and disadvantages. Based on this, we provide an improved algorithm to combine the advantages of these methods. To prove the effects, some real data were adopted for the calculation and simulation which were also shown in this paper.