An Interaction-Aware Predictive Control Framework with Adaptive Gap Regulation in Connected and Automated Vehicle Platoon

The deployment of Connected and Automated Vehicles (CAVs) holds great promise for improving traffic safety, efficiency, and sustainability. Among CAV applications, vehicle platooning enables cooperative driving with reduced space gaps and increased capacity. However, in mixed traffic with Human-Driven Vehicles (HDVs), unpredictable behaviors and communication delays introduce significant challenges for safe and efficient platoon control. To address these issues, we propose IAG-MPC, an interaction-aware model predictive control framework for CAV platoons. It integrates a graph-based trajectory prediction module and a dynamic space gap regulation strategy to handle communication delays. This framework allows CAV platoons to adaptively adjust inter-vehicle space gaps in response to the motion of surrounding vehicles while maintaining safety and string stability. Experiments on the HighD dataset demonstrate that IAG-MPC outperforms baseline controllers in terms of space gap accuracy, speed stability, and overall efficiency under real-world uncertainties.