<p>Intelligent agents operating in uncertain environments must quickly establish mobile decision-making, and the multi-agent path-finding is increasingly focused on improving adaptability to various scenarios. When multiple agents collaborate to search for dynamic targets in uncertain settings, they encounter challenges related to conflicts. A new online coordination layer has been proposed to address the conflicts that arise during the online planning of distributed systems. This layer utilizes an improved conflict-based search algorithm, which outlines the main conflicts encountered in online multi-agent pathfinding and resolves them using a priority decision algorithm grounded in a non-cooperative game model, alongside an improved algorithm that integrates multi-objective decision analysis with enhanced conflict-based search. This comprehensive strategy aims to enhance the adaptability of multi-agent systems to changes in targets and to improve their navigation in uncertain environments, even when they encounter inaccurate information. Finally, numerical examples are conducted across various scenarios to evaluate the feasibility of the proposed algorithm. The proposed algorithm was compared with several global planning algorithms to validate its effectiveness and demonstrate its applicability in online environments. The experimental results indicate that the proposed algorithm can effectively resolve conflicts in online multi-agent path finding, enabling the system to navigate successfully and capture targets.</p>