Compliance Factor-Based Artificial Potential Fields for Collision-Aware UAV Swarm Tracking

Abstract

The deployment of Unmanned Aerial Vehicle (UAV) swarms poses a major technical challenge for modern surveillance, search-and-rescue, and industrial inspection missions. Unlike traditional monolithic architectures that rely on a single central supercomputer, this paper proposes a fully decentralized control architecture. The system is developed using the Robot Operating System 2 (ROS2) framework, leveraging its Data Distribution Service (DDS) protocol to ensure asynchronous and robust node communication without a single point of failure. The objective is to coordinate a fleet of five drones to navigate autonomously in a complex environment, avoid collisions, and track a moving target. Using a PID controller for low-level stabilization and an Artificial Potential Field (APF) approach enhanced with a dynamic compliance factor, the swarm can dynamically elect a tracker and maintain a tactical formation. The results demonstrate the proposed method’s potential for high-speed, real-time obstacle avoidance in a simulation environment while maintaining decentralized communication among the UAVs.