针对一类具有不确定性的动态关联大系统，提出了一种基于分散重叠观测器设计的专家分散鲁棒控制算法．专家系统利用不同重叠观测器所估计的状态之间的误差来确定故障的位置，由此选择合适控制器和观测器的增益来保证整个系统的稳定性，并使从干扰输入到被控输出的闭环传递函数的Ｈ∞范数小于１个给定的常数 Aiming at a class of uncertain large-scale dynamic association system, an expert-based decentralized robust control algorithm based on decentralized overlapping observer design is proposed. The expert system uses the errors between the states estimated by different overlapping observers to determine the location of the fault and thereby selects the appropriate controller and observer gain to ensure the stability of the overall system and to allow the input from the disturbance to the controlled output The H∞ norm of the closed loop transfer function is less than 1 given constant