In this paper, a robust fault diagnosis method is proposed to estimate the states and faults simultaneously for a special class of Lipschitz nonlinear systems. In the model of system, fault is considered as a linear and additive function and disturbance as a nonlinear function coupled with the system states. An augmented system is constructed by forming a vector composed of states and faults, and a Luenberger observer is designed for fault estimation. The nonlinear function of state-coupled disturbance is replaced by a Lipschitz matrix. In order to reduce the conservatism of the problem, this matrix depends on both the states and disturbances. This approach attempts to attenuate the effects of the state-coupled disturbances using the dissipativity theory such that the existing problem finally leads to solving convex optimization problems. The necessary conditions for the existence of such an observer are expressed. Finally the performance of the proposed method is simulated on a robot, and the results indicate good performance of the proposed method.

中文翻译：

---

具有状态耦合扰动的 Lipschitz 非线性系统的基于观测器的鲁棒故障估计：耗散性方法

在本文中，提出了一种鲁棒故障诊断方法来同时估计一类特殊的 Lipschitz 非线性系统的状态和故障。在系统模型中，故障被认为是一个线性的加性函数，扰动被认为是一个与系统状态耦合的非线性函数。通过形成由状态和故障组成的向量来构建增强系统，并设计了 Luenberger 观测器用于故障估计。状态耦合扰动的非线性函数由 Lipschitz 矩阵代替。为了降低问题的保守性，这个矩阵同时取决于状态和扰动。这种方法尝试使用耗散理论来减弱状态耦合扰动的影响，从而使现有问题最终导致解决凸优化问题。表达了这种观察者存在的必要条件。最后在机器人上对所提方法的性能进行了仿真，结果表明所提方法的性能良好。