基于逆系统迭代学习观测器的故障调节方法

针对一类满足Lipschitz条件的多输入多输出非线性可逆系统执行器故障问题,提出了一种基于迭代学习观测器的逆系统内模故障调节方法。引入PD型迭代学习策略,设计了迭代学习故障诊断观测器,用于对执行器未知时变故障进行快速、准确估计。根据故障估计值,结合逆系统方法对逆模型进行补偿,使得补偿后的逆模型与非线性被控对象串联仍为伪线性系统;再结合内模控制实现了伪线性系统的容错控制。最后,通过仿真算例验证了该方案的有效性。

Class method of fault accommodation based on inverse systemiterative learning observer

This paper proposed a designing method of fault accommodation based on iterative learning observer and combined both inverse system method and internal model control for a class of multi-input multi-output nonlinear invertible systems which satisfy Lipschitz conditions with actuator failure problem. Introducing a PD-type learning strategy, it designed an iterative learning observer to estimate the actuator fault rapidly and exactly. Then provided fault estimate for compensating inverse system model to keep the nonlinear system plant connected by inverse model pseudo-linearization, and, combined internal model control to carry fault tolerant control of the pseudo-linearization system. Finally, simulation results show the effectiveness of the scheme.