基于未知输入观测器的非线性广义系统执行器故障检测和重构

针对一类含有未知扰动广义非线性系统的执行器故障，本文提出一种重构算法。首先设计未知输入观测器对干扰鲁棒，作为故障检测观测器。检测到发生故障后，通过提出含有误差比例项和积分项的故障估计算法，形成自适应观测器，实现准确快速地估计故障，同时估计状态变量。根据李雅普诺夫稳定理论给出估计误差一致最终有界的充分条件。最后仿真验证该类观测器和重构算法的有效性。

Actuator Fault Detection and Reconstruction Based on Unknown Input Observer in Nonlinear Generalized System

An adaptive fault reconstruction algorithm is proposed to deal with the actuator fault for a class of generalized nonlinear system with unknown disturbance. Firstly, the unknown input observer as a detection observer is designed to achieve robust to unknown disturbance. After detecting the fault, an adaptive observer is developed by using an adaptive law which contains the proportional and integral terms of error to simultaneously estimate the fault and states rapidly and accurately. Then, the sufficient conditions are presented to ensure the estimation error to be uniformly bounded according to Lyapunov stable theory. Finally, an example is given to show the effectiveness of the developed observers and fault diagnosis algorithms.