Accurate and real- time prediction of traffic speed is an essential part of intelligent traffic control and construction of smarttransportation system. However, the existing prediction methods cannot mine the potential spatio - temporal correlation in the dataaccurately. In order to further mine the spatio-temporal characteristics of data?and improve the prediction accuracy,a GRU-GCN-RDropcombined model based on gated recurrent unit (GRU) , graph convolutional network ( GCN) and regularized dropout ( R - Drop) isdesigned. The GCN is used to learn complex topological structures to capture spatial dependence and the gated recurrent unit is used tolearn dynamic changes of traffic data to capture temporal dependence. After the combination of GCN and GRU,
the R-Drop method isused to improve the generalization ability of the model. SZ-taxi data set is taken as an example for prediction and analysis. GRU-GCN-RDrop model predicts the future traffic speed in 15 minutes,
30 minutes,45 minutes and 60 minutes, and obtains the correspondingRMSE,MAE,Accuracy,R2 and Var. Compared with GCN and GRU models,GRU-GCN-RDrop model effectively solves the problemof rapid accumulation of errors. Compared with other benchmark models,GRU-GCN-RDrop model is superior in feature extraction andinterpretation of traffic speed series. Compared with T - GCN model and ST - AGTCN model,
GRU - GCN - RDrop model has strongergeneralization ability. It is showed that the timeseries of traffic speed predicted by GRU-GCN-RDrop model has high accuracy and stability.