Dynamic Trajectory Planning for Emergency Vehicle Clearance at Traffic Intersections Using Model Predictive Control

This paper presents an innovative approach for optimizing the clearance of emergency vehicles at traffic intersections by employing Model Predictive Control (MPC) in conjunction with reference trajectory generation. The algorithm operates in two distinct phases: offline static map generation and online dynamic trajectory planning. In the offline phase, the algorithm constructs a static map of the intersection, approximating the drivable area with a polytope covering. In the online phase, the algorithm continuously gathers real-time data on the positions of all vehicles present at the intersection. Based on mixed-integer programming techniques, our algorithm dynamically generates reference trajectories for each vehicle, including the emergency vehicle to facilitate the fastest possible passage for the emergency vehicle to its target location while ensuring the safe clearance of the path ahead. We demonstrate the feasibility and effectiveness of our model predictive control-based algorithm in enhancing the response time of emergency vehicles and minimizing intersection congestion, ultimately contributing to the improvement of urban safety and emergency response services.