IP 3 C: Integrated Path Planning and Predictive Control for UAV Swarm Navigation in Cluttered Environment

This paper presents a path planning methodology specifically designed for the coordinated navigation of Unmanned Aerial Vehicles (UAVs) exhibiting swarming behavior in urban environments. Coordinating a team of UAVs in dynamic environments necessitates the real-time generation of conflict-free trajectories that account for both moving obstacles and the behavior of neighboring agents. To address this challenge hierarchically, the proposed framework integrates a global planner based on Rapidly-exploring Random Trees Star (RRT*) at the upper layer with a Distributed Model Predictive Control (DMPC) algorithm at the lower layer. The RRT* planner generates reference trajectories in a static environment, providing high-level guidance to the swarm. These trajectories are then refined and tracked by each UAV in the swarm with a predictive mechanism through the DMPC, which performs local collision avoidance and trajectory optimization. This novel strategy ensures that each UAV follows its designated path with minimal deviation while safely avoiding obstacles in a cluttered urban environment and maintaining minimal inter-agent separation allowed within the swarm. Numerical simulations demonstrate the effectiveness of the proposed approach in ensuring safe and coordinated multi-agent navigation. The results highlight the framework's potential for real-world deployment in time-critical urban missions involving large UAV teams.