Zhao, Xia (Beijing University of Civil Engineering and Architecture), Menglin, Wu (Beijing University of Civil Engineering and Architecture), Li, Zhihong (Beijing University of Civil Engineering and Architecture), Qin, Yimeng (Beijing University of Civil Engineering and Architecture), Peilin, lv (Beijing University of Civil Engineering and Architecture)

Adaptive Clustering of Trucks with High-Order Mobility Correlations Using Enhanced Hypergraph Convolutional Networks

Scheduled for presentation during the Video Session "On-Demand Video Presentations" (VP-VP), Saturday, November 22, 2025, 08:00−18:00, On-Demand Platform

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on April 2, 2026

Abstract

We propose an enhanced hypergraph convolutional network (EHGCN) to address precise truck clustering through advanced mobility correlation analysis. Existing approaches fail to sufficiently investigate latent travel patterns beneath observable features or comprehensively model high-order inter-truck relationships, leading to suboptimal grouping accuracy. The proposed framework generates a multidimensional feature representation characterizing vehicular movements across temporal, spatial, and attribute dimensions, subsequently constructing hyperedges to encapsulate higher-order dependencies within localized clusters. The architecture operates a node-level autoencoder that deciphers implicit mobility patterns from explicit travel features via nonlinear feature transformation, while structural-level hypergraph convolutions capture intricate group-wise correlations through neighborhood aggregation. An adaptive attention module fuses these representations, optimizing feature relevance weighting for enhanced discriminative capability in cluster identification. Comprehensive evaluations across three datasets confirm the model’s efficacy through comparative and ablation studies. Experimental results demonstrate EHGCN’s superiority over conventional GCN variants, achieving improvement in clustering purity and robust performance. The framework successfully categorizes six clusters of trucks, including regional distribution fleets, cross-regional logistics units, and specialized commodity transporters. The derived classifications can offer operational insights for optimizing freight management through route rationalization, fleet coordination, and service customization, thus enhancing the development of intelligent transportation systems.