描述Agent间的交互关系 ,提出了一种基于域的多Agent协作策略 .该策略限定信息传播和共享的范围 ,实现Agent间负载的动态调整 ,它遵循就近原则 ,在局部范围内寻找最优的合作者 ,协作完成子计划 ,从而平衡负载 ,降低开销 ,扩大分布式计算环境的处理能力 .为MADCE提供了一种有效的协作策略 ,并给出了算法描述、评价及例示 . Describes the interaction between Agents and proposes a domain-based multi-agent collaboration strategy that limits the scope of information dissemination and sharing and realizes the dynamic adjustment of load among Agents. It follows the principle of proximity and looks for the optimal Collaborators collaborate on sub-plans to balance load, reduce overhead and expand the processing power of distributed computing environments, provide an effective collaboration strategy for MADCE, and provide algorithm descriptions, evaluations and examples.