In order to improve the prediction accuracy of expressway traffic flow, combined with S-G filter, bidirectional long short-term memory(BiLSTM), gated recurrent unit(GRU) and other methods, a BiLSTM-GRU combined model based on S-G filter was proposed to predict the short-term traffic flow of expressways. First, the S-G filter was employed to reasonably remove noise and abnormal fluctuations from the data, addressing the issue of models struggling to capture the high volatility of traffic flow. Second, BiLSTM was utilized to leverage its bidirectional advantage in capturing long-term dependencies and temporal patterns in traffic flow data. Third, GRU was employed to automatically learn features, efficiently extracting and transforming the output features of BiLSTM to produce predicted values. Finally, the mean absolute error (MAE), mean square error (MSE), and mean absolute percentage error (MAPE) were used to evaluate and analyze the prediction results. The case analysis selects the traffic flow data of the 104 km main line bayonet Shenyang direction of the Beijing-Harbin Expressway in Tianjin (Tangshan section), and analyzes the prediction performance of the combined model. The results show that the BiLSTM-GRU combined model based on S-G filter improves MSE by 0.75 to 51.28, MAE by 0.14 to 3.82, and MAPE by 0.95% to 20.84% compared with the traditional single model and other combined models. It can be seen that the BiLSTM-GRU combined model based on S-G filter can reasonably remove noise and abnormal fluctuations in the data, improve the prediction accuracy of short-term traffic flow on expressways, and provide a theoretical basis for traffic flow control on expressways.