Knowledge fusion enhanced graph neural network for traffic flow prediction

Article Engineering, Civil

Fine-Grained Vessel Traffic Flow Prediction With a Spatio-Temporal Multigraph Convolutional Network

Maohan Liang et al.

Summary: Accurate and robust vessel traffic flow prediction is crucial in maritime intelligent transportation systems. To achieve this, a graph-driven neural network is proposed, which utilizes the maritime traffic network and extracts feature points from vessel positioning data. The method demonstrates superior performance in accuracy and robustness.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2022)

Add to Collection

Article Engineering, Civil

AARGNN: An Attentive Attributed Recurrent Graph Neural Network for Traffic Flow Prediction Considering Multiple Dynamic Factors

Ling Chen et al.

Summary: This paper proposes a graph neural network-based method for traffic flow prediction, considering multiple dynamic factors. By combining multi-source urban data and attention mechanism, the method can predict short-term traffic flow more accurately.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2022)

Add to Collection

Article Engineering, Civil

Metro Traffic Flow Prediction via Knowledge Graph and Spatiotemporal Graph Neural Network

Shun Wang et al.

Summary: Existing traffic flow prediction methods often overlook the interference of various external factors in addition to the spatiotemporal characteristics of traffic flow. In this study, we propose a novel metro traffic flow prediction method (KGR-STGNN) based on knowledge graph representation learning, which incorporates the influence of external factors into the prediction model through learning the influence representation from a traffic knowledge graph. Experimental results demonstrate the effectiveness of our proposed model.

JOURNAL OF ADVANCED TRANSPORTATION (2022)

Add to Collection

Article Engineering, Civil

Metro Passenger Flow Prediction via Dynamic Hypergraph Convolution Networks

Jingcheng Wang et al.

Summary: This paper proposes a dynamic spatio-temporal hypergraph neural network to forecast metro passenger flow, utilizing hypergraph convolution and spatio-temporal blocks to extract spatial and temporal features for node-level prediction. Experiments on historical datasets from Beijing and Hangzhou validate the effectiveness and superior performance of the proposed method in prediction accuracy compared with the state-of-the-arts.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2021)

Add to Collection

Article Engineering, Civil