Short-term traffic flow prediction method based on grey adaptive particle swarm support vector machine

Abstract

In order to further improve the accuracy of the short-term traffic flow prediction, a combination of short-term traffic flow prediction model had been proposed by analyzing the characteristics of grey model, adaptive particle swarm optimization (PSO) algorithm and support vector machine (SVM) model. First, use the grey model to accumulate the original traffic flow data, weaken the randomness of the traffic flow data sequence, then optimize the support vector machine model based on adaptive particle swarm optimization algorithm and realize short-term traffic flow prediction, finally, get the final predicted value table by grey mode. The model was verified based on the traffic flow data of the major road in Changchun and the experimental result showed the proposed model was effective and feasible.