针对系统模型的不确定性、未知输入扰动和非线性特性,提出一类非线性系统参数估计的故障诊断算法.构造系统故障诊断观测器,采用Lyapunov稳定性定理验证观测器的稳定性,通过Barbalat引理证明满足故障诊断观测器为渐近稳定的表征故障参数的参数估计,并总结了设计算法流程.仿真结果表明,所提出算法具有快速收敛性,对一类非线性系统诊断效果较好. Aiming at the uncertainties of system model, unknown input disturbance and nonlinear characteristics, a fault diagnosis algorithm for nonlinear system parameter estimation is proposed. The system fault diagnosis observer is constructed, and the stability of observer is verified by Lyapunov stability theorem. By Barbalat Lemmatizing proved that the fault diagnosis observer is asymptotically stable parameter estimation of fault parameters and summarized the flow of the design algorithm.The simulation results show that the proposed algorithm has fast convergence and good performance for a class of nonlinear systems.