A Spatial and Temporal Combination Model for Traffic Flow: A Case Study of Beijing Expressway

Short-term traffic flow prediction is playing an important role in the intelligent transportation system. However, exploring high-precision and efficient prediction methods is still a challenge. To capture the spatiotemporal characteristics of traffic flow and accurately perceive the traffic state, a spatial and temporal combination (STC) model was proposed. The radial basis function neural network (RBFNN) was used to capture the spatial characteristics of traffic flow, while the clockwork recurrent neural network (CWRNN) was utilized to predict the temporal characteristics. The prediction accuracy of the model can be further improved by the result fusion based on the spatiotemporal feature prediction model. To verify the accuracy and robustness of the algorithm, the Beijing 3rd Ring Road speed data are used to compare with other models. The results show that the accuracy of the STC algorithm is better than benchmark prediction models at different service levels.