A traffic flow prediction algorithm based on adaptive particle filter

The online traffic flow prediction is an important part of the road-traffic management system. If the traffic flow prediction real time capability is not strong enough, the prediction outcomes will become uncertain. The adaptive particle filter algorithm suggested in this paper is based on confidence level. And this algorithm can adaptively adjust the number of the particles according to the state of the particles and reduce the quantity of the particle filter algorithm calculation so as to improve the real time capability, on the condition of guaranteeing the algorithm precision. The experiment results have verified the effectiveness of the method.