Real-Time Path-Reconfigurable Coverage Planning for Multi-UAV Missions Over Disjoint Areas
Cai Luo, Jintao Guo, Zixuan Liu, LEI LIU, Chunbo Luo
AI summary
Problem
Multi-UAV missions across geographically separated regions face high coordination costs and lack robust mechanisms for handling in-flight failures without disrupting coverage continuity or energy constraints.
Approach
The authors introduce GRIT-M, which uses a composite transition cost to optimize initial region sequencing and task allocation, then applies a region-priority greedy repair and region-preserving Tabu search to rapidly replan paths after UAV failures.
Key results
- Reduces inter-region flights by up to 27% compared to baselines
- Maintains workload balance with a coefficient of variation below 7%
- Achieves over 91% replanning success rate with average repair times under 2.1 seconds
- Scales effectively to 12-UAV teams while preserving coverage coherence
Why it matters
Provides a computationally efficient, fault-tolerant planning framework essential for time-critical aerial survey and search-and-rescue operations over fragmented terrains.
Abstract
Multi-uncrewed aerial vehicle (UAV) cooperative cov- erage missions across geographically separated regions face signif- icant challenges due to potential UAV failures during mission exe- cution. To address the challenges of efficient multi-region coverage and dynamic failure handling, this paper presents a novel real- time path-reconfigurable coverage-planning algorithm for multi- UAV coverage path planning across separated geographical areas with real-time path reconfiguration capabilities. The proposed approach, GRIT-M (Greedy Repair Initializes Tabu search for Multiple regions), extends the GRIT algorithm to efficiently handle both initial planning and online path repair when UAVs fail dur- ing mission execution. Unlike existing methods that either focus on single-region coverage or lack failure handling mechanisms, GRIT-M incorporates region-specific knowledge through three key innovations: (1) a composite transition cost function that opti- mizes inter-region movements, (2) a region-priority greedy repair mechanism that minimizes transition costs during replanning, and (3) a region-preserving Tabu search that maintains area cover- age coherence while balancing workload. Extensive simulations demonstrate GRIT-M’s scalability and robustness against opera- tional constraints. The algorithm reduces inter-region flights by up to 27% compared to baselines and maintains excellent workload balance (CV < 7%). Crucially, it shows effective scalability with larger teams (up to 12 UAVs), maintaining a > 91% replanning success rate with average repair times under 2.1s. Its resilience to variations in battery capacity, sweep patterns, and initial UAV distributions confirms its suitability for time-critical aerial surveys over disjoint areas.