Large-Scale Battery-Conscious Collision-Free Path-Finding for Sustainable and Autonomous Multi-AGV Mobility Control

This paper proposes a novel battery-conscious collision-free sustainable path-finding (BCSP) algorithm for large-scale multi-agent systems (MAS). The framework rigorously integrates battery management with reservation-based collision avoidance with local observation. Agent states, charging station availability, and spatiotemporal reservations are explicitly formulated to ensure safe navigation. An octile distance-based heuristic search strategy minimizes redundant node expansions and reduces computational overhead, utilizing adaptive waiting and local reservation updates. Furthermore, performance evaluations demonstrate the superiority of the proposed algorithm in terms of battery efficiency and collision avoidance, outperforming existing benchmark algorithms. Moreover, extensive experiments in highly congested and dynamic environments even reveal marked improvements compared to benchmarks. The performance and robustness of the algorithm underscore its potential for real-world applications in autonomous navigation, smart factories, and urban mobility.