Short-term traffic flow prediction during peak hours on highways based on GBDT

Traffic congestion during highway peak hours, caused by concentrated traffic flow, underscores the importance of accurate short-term traffic prediction for optimizing intelligent transportation systems. However, traditional prediction models struggle to capture the unique characteristics of short-term traffic flow during peak periods. This paper proposes a short-term traffic flow prediction model based on the Gradient Boosting Decision Tree (GBDT) algorithm. By identifying peak hours through covariance analysis to construct training datasets, the model’s effectiveness is enhanced. Experiments on highway data from a Chinese province demonstrate the superior performance of the proposed prediction model in traffic flow prediction.