基于交通小区短时交通生成预测的城市道路网短时动态交通预测研究OACSTPCD

The difference of land use nature around urban roads leads to the difference of short-time occurrence and attraction traffic volume,which directly affects the traffic flow on the urban road network.This paper makes short-time prediction of occur-rence and attraction traffic volume of each plot in the urban road network,and allocates the predicted traffic volume to the urban road network through dynamic traffic,so as to realize short-time dynamic traffic prediction of the urban road network.The single BP neural network model is used as the comparison model,and the model is trained and parameter calibrated,and the prediction effect of the model is tested.The experimental results show that the prediction accuracy of the proposed prediction method can be improved by up to 84.28%compared with that of BP neural network.