从Agent理论出发 ,构造了一种面向Agent的智能大系统。系统由一个协调控制子系统和多个Agent组成的多Agent系统 ;协调控制子系统为系统的全局数据库和监控、协调、优化中心。Agent之间的协作有任务分担和结果共享两种方式 ,它们之间的冲突是通过多轮协商法来解决的。大系统的优化、控制就是通过上述多Agent系统实施 ,以达到对大系统进行动态优化、控制和模型的自适应的目的。 Based on Agent theory, an agent-oriented intelligent big system is constructed. The system consists of a multi-agent system consisting of a coordinating control subsystem and a plurality of agents. The coordinating control subsystem is the global database of the system and the monitoring, coordination and optimization center. There are two ways to share the tasks and share the results among the agents. The conflicts between them are solved through many rounds of negotiation. The optimization and control of the large system is implemented through the above multi-agent system in order to achieve the purpose of dynamic optimization, control and model adaptation of the large system.