Space matching fusion model for arterial speed estimation in ITS

To improve the accuracy of arterial mean speed estimation through data fusion in road traffic, this paper presented a speed estimation method based on space-matching fusion model. In the method, an urban road model is proposed, which divided road into several equal length segments. The loop detector data is matched to each segment by the adjusted coefficient. Because of the strongly complementary between loop vehicle detector data and probe vehicle data in space, a weighted fusion model is proposed. The model is used to calculate arterial mean speed by link mean speed of each segment. In the weighted fusion model, Newton method is used as training method to get the weights. The simulation results show that this method is effective and reliable to improve the accuracy of arterial link mean speed, and the computation is reduced owing to relatively simple training process.