Accurate traffic flow prediction can provide travelers with real-time and reliable traffic information,avoiding the risk of traffic accidents caused by road congestion.Therefore,a precise prediction method for highway traffic flow based on MC-STGNN model is proposed.Firstly,a millimeter wave radar system is established to collect highway traffic flow data;Secondly,fill in missing values in the collected data to improve the accuracy of data collection;finally,an MC-STGNN model is established to divide the collected traffic flow data into periodic time series,and the dynamic correlations between highway nodes are obtained by encoding their temporal and spatial positions.By setting up a multi head self attention mechanism,the temporal correlations between traffic flow data are analyzed,and a graph generator combined with gate graph convolution is established to predict highway traffic flow.The experimental results show that the proposed method has high data acquisition accuracy and prediction accuracy.
关键词
毫米波雷达/高速公路/缺失值填补/交通流量/预测模型
Key words
millimeter wave radar/expressway/fill in missing values/traffic flow/predictive model
万方数据