A Digital Twin-Based Approach for Dynamic Interactive Information Management at Intersections

This paper addresses key challenges in current intelligent intersection management, including coarse information granularity, static control strategies, and weak system coordination, by proposing a comprehensive lane-level cooperative management and control method based on digital twin technology. By constructing a high-fidelity, low-latency digital twin of the intersection, a systematic architecture is established comprising four core models: dynamic lane function modeling, spatiotemporal traffic corridor generation, dynamic spatiotemporal conflict analysis, and coordinated decision-making for priority vehicle passage. This architecture innovatively transforms abstract signal phases into concrete spatiotemporal traffic corridors, achieving a digital description of the right-of-way. Through multi-source data fusion and model coupling, precise, lane-level and vehicle-level dynamic control instructions are generated. Research results demonstrate that the proposed method effectively enhances intersection traffic efficiency, significantly improves safety via dynamic conflict probability quantification, and provides superior global decision-making support for high-level automated vehicles beyond onboard perception limitations. The systematic solution presented herein offers a crucial technical pathway towards realizing a vehicle-road-cloud integrated next-generation intelligent transportation system.