Battery-Aware Dynamic Leader?Follower Switching for Cooperative UAV Systems = 배터리 인식 기반 동적 리더?팔로워 스위칭 UAV 협력 시스템

다국어 초록 (Multilingual Abstract)

This thesis presents the design and implementation of a battery-aware leader–follower UAV cooperation system built on the ROS 2 distributed framework. Conventional multi-UAV cooperation techniques rely on a static leader, which makes the entire formation vulnerable to early mission termination when the leader experiences battery depletion, sensor degradation, or communication instability. To address these limitations, this study proposes a dynamic role-switching architecture that continuously evaluates each UAV’s energy status and communication quality to adaptively reassign leadership.
The proposed system integrates four modular components—role_manager, battery_role_switcher, leader_traj_broadcaster, and guide_adapter—and supports seamless switching using a soft-handover mechanism that preserves trajectory continuity during leader transitions. In addition, the architecture supports multi-leader extension and partial leaderless operation under intermittent communication, enabling robustness in complex or constrained mission environments.
The system is implemented in the ego_swarm_coop package and integrated with the ego-planner (ros2_version) trajectory planning framework. Two representative scenarios—static leader operation and dynamic energy-based switching—were evaluated. Experimental results show that the proposed dynamic strategy reduces total SoC consumption by approximately 17%, decreases energy imbalance by 78%, and improves distance-normalized energy efficiency. These results demonstrate that energy-aware cooperative behavior significantly enhances endurance, fairness, and mission continuity in multi-UAV operations.