中文摘要：

为提高复杂网络环境和任务需求下器材保障的智能化水平，提出了一个数据驱动的多Agent器材保障决策支持系统DSS（decision support system），系统中的数据、模型以N．Agent可分布在不同的网络节点上，高层Agent可由粒度更小的子Agent或Agent部件组合而成；系统中的各类智能Agent相互协作，共同完成异构数据集成、多维数据集构造、器材保障模型挖掘、复杂规划问题求解，以及保障方案自动生成等任务，从而有效降低了系统开发的难度，提高了Agent协作的灵活性。研究结果表明，该系统架构能够显著提升器材保障决策方案的质量和效率。

英文摘要：

A data-driven multi-agent decision support systems （DSS） was proposed to improve intelligent and high-performance equipment support. System data ,problem models, and agents could be distributed at different nodes in network, and the agents could be further composed of sub-agents and of agent components. System agents cooperated to integrate heterogeneous data sources, construct multi-dimensional data cubes, mine valuable equipment support models, solve large-scale constraint satisfaction problems and to generate a variety of support plans automatically and intelligently, and in consequence decreased the difficulty in system development and increased the interactive flexibility. The result demonstrates that the distributed, component-based architecture can significantly improve the quality and effectiveness of equipment decision-support solutions.