针对非线性不确定系统的故障诊断问题,提出了一种基于主导子系统的自适应模糊观测器设计方法。利用主导子系统的观测器增益矩阵进行模糊观测器的设计,并推导出了系统状态估计误差有界的条件以及故障估计所要满足的自适应算法。另外,考虑到系统带有建模不确定,对故障检测的鲁棒性和灵敏度进行了分析。给出的仿真示例说明所提出的方法是有效的。 Aimed at the fault diagnosis of nonlinear uncertain systems, a design method of adaptive fuzzy observer based on dominating subsystems is proposed. The observer gain matrix of the dominant subsystem is used to design the fuzzy observer. The condition of the bounded state estimation error and the adaptive algorithm to estimate the fault estimate are deduced. In addition, considering the system modeling uncertainty, the robustness and sensitivity of fault detection are analyzed. The given simulation example shows that the proposed method is effective.