Planning smart link-based geofences for urban sustainability via speed limit management

Urban transportation faces severe challenges, including congestion, air pollution, and increased fuel consumption. With the widespread deployment of road monitoring equipment and the emergence of connected automated vehicles, precise speed management has become feasible. This paper combines link-based geofencing technology with speed management and proposes a smart link-based geofence planning (SLGP) approach based on deep reinforcement learning. We designed a sustainability-oriented reward function, enabling SLGP to simultaneously optimize traffic efficiency and reduce emission and fuel consumption when planning speed limits. Furthermore, we built a digital twin simulation platform that integrates multiple functional modules and scenarios for comprehensively evaluating the effectiveness of the SLGP. The results indicate that the SLGP significantly enhances urban transportation sustainability across various scenarios, particularly under high-volume scenarios, where it effectively reduces emissions and fuel consumption. This study provides robust support for leveraging digital twin and artificial intelligence technologies to empower traffic management and promote sustainability.