.png){kind=icon}
.png){kind=icon}
Return
Heterogeneous Multi-agent Coverage Control Through Adaptive Weighting and Corrective Potential Function
.png){kind=icon}
PRE
AI
C
P
I
H
DOI：10.1007/s40313-026-01268-8.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper presents an integrated adaptive coverage control framework for multi-agent systems that couples distributed weight learning with corrective potential fields, enabling robust coordination in heterogeneous, dynamic, and obstacle-rich environments. Unlike existing approaches that address adaptive coverage and obstacle avoidance separately, we develop a unified framework providing formal Lyapunov-based stability guarantees for simultaneous weight adaptation and navigation dynamics. Each agent dynamically adjusts its power diagram weight based on real-time sensing quality, environmental complexity, and resource availability; simultaneously, a corrective artificial potential field eliminates spurious local minima while enabling collision-free navigation through obstacle-rich domains. Comprehensive theoretical analysis establishes that weight and navigation dynamics are mutually compatible and can benefit each other: adaptive weighting facilitates faster convergence to locally optimal configurations. Numerical validation shows promising results: a 99.08% improvement in coverage cost for coordinated 60-agent teams navigating 30 obstacles; graceful degradation limiting coverage loss to 12% under catastrophic failures (compared to 35.6% for non-adaptive methods); and sub-linear scaling to 40-agent teams. These results validate the framework's practical applicability to large-scale multi-agent systems under realistic constraints including agent capability degradation, time-varying importance distributions, and complex obstacle environments.
Keywords:
Coverage control
Distributed algorithms
Multi-agent systems
Power diagrams
Potential field methods
Adaptive learning
Journal
J
IF:
1.3
Papers:
92
Citations:
1.1K
Organization
Cited Papers