Make More Connections: Urban Traffic Flow Forecasting with Spatiotemporal Adaptive Gated Graph Convolution Network

Urban traffic flow forecasting is a critical issue in intelligent transportation systems. Due to the complexity and uncertainty of urban road conditions, how to capture the dynamic spatiotemporal correlation and make accurate predictions is very challenging. In most of existing works, urban road network is often modeled as a fixed graph based on local proximity. However, such modeling is not sufficient to describe the dynamics of the road network and capture the global contextual information. In this paper, we consider constructing the road network as a dynamic weighted graph through attention mechanism. Furthermore, we propose to seek both spatial neighbors and semantic neighbors to make more connections between road nodes. We propose a novel Spatiotemporal Adaptive Gated Graph Convolution Network ( STAG-GCN ) to predict traffic conditions for several time steps ahead. STAG-GCN mainly consists of two major components: (1) multivariate self-attention Temporal Convolution Network ( TCN ) is utilized to capture local and long-range temporal dependencies across recent, daily-periodic and weekly-periodic observations; (2) mix-hop AG-GCN extracts selective spatial and semantic dependencies within multi-layer stacking through adaptive graph gating mechanism and mix-hop propagation mechanism. The output of different components are weighted fused to generate the final prediction results. Extensive experiments on two real-world large scale urban traffic dataset have verified the effectiveness, and the multi-step forecasting performance of our proposed models outperforms the state-of-the-art baselines.

A Grade Identification Method of Critical Node in Urban Road Network Based on Multi-Attribute Evaluation Correction

Accurately identifying the key nodes of the road network and focusing on its management and control is an important means to improve the robustness and invulnerability of the road network. In this paper, a classification and identification method of key nodes in urban road networks based on multi-attribute evaluation and modification was proposed. Firstly, the emergency function guarantee grade of road network nodes was divided by comprehensively considering the importance of road network nodes, the consequences of failure, and the degree of difficulty of recovery. The evaluation indexes were selected according to the local attributes, global attributes, and functional attributes of the road network topology. The spatial distribution patterns of the evaluation indexes of the nodes were analyzed. The dynamic classification method was used to cluster the attributes of the road network nodes, and the TOPSIS method was used to comprehensively evaluate the importance ranking of the road network nodes. Attribute clustering of road network nodes by dynamic classification method (DT) and the TOPSIS method was used to comprehensively evaluate the ranking of the importance of road network nodes. Then, combined with the modification of the comprehensive evaluation and ranking of the importance of the road network nodes, the emergency function support classification results of the road network nodes were obtained. Finally, the method was applied to the road network within the second Ring Road of Beijing. It was compared with the clustering method of self-organizing competitive neural networks. The results show that this method can identify the key nodes of the road network more accurately. The first-grade key nodes are all located at the more important intersections on expressways and trunk roads. The spatial distribution pattern shows a “center-edge” pattern, and the important traffic corridors of the road network show a “five vertical and five horizontal” pattern.

An Analysis of the Catastrophe Model and Catastrophe Characteristics of Traffic Flow Based on Cusp Catastrophe Theory

—The state of urban road traffic flow shows discontinuity and jumping phenomenon in the process of running. There was a data gap in the collected traffic flow data. Through the data analysis, it was found that the traffic flow state had the characteristics of multimode, mutation, inaccessibility, divergence and hysteresis, which were similar to the mutation characteristics of the basic model of catastrophe theory when the system state changed. The cusp catastrophe model of traffic flow based on traffic wave theory was established by analyzing the movement process of traffic flow. In this model, the traffic density was taken as the state variable, and traffic flow and wave speed were taken as the control variable. Referring to the basic idea of catastrophe theory, the solution method of the model was given, and the structural stability of the traffic flow state was analyzed. Through the critical equilibrium surface equation, the stability of the extreme value of the system potential function can be analyzed, and the bifurcation set equation when the traffic flow state changed can be obtained, which can be used to determine the critical range of the structural stability of the system. This paper discussed and analyzed the changing trend and constraint relationship among the wave speed, traffic density and traffic flow when the traffic flow state changed suddenly in different running environments. The analysis results were consistent with the actual road traffic flow state. A case was given, and the results showed that the cusp catastrophe model could describe the relationship among the three parameters of traffic flow from three-dimensional space, and could effectively analyze the internal relationship of the parameters when the traffic flow state changed. The validity of the model and analysis method was verified. The goal of this paper is to provide an analysis method for the judgment of urban road traffic state.