基于级联观测器的多输入多输出非线性系统的故障诊断

将块观测方法应用于非线性系统的故障检测和分离.首先给出了非线性系统的块观测形式,针对传感器故障和执行器故障对非线性系统进行分块,得到了带有故障系统的块观测器形式.利用滑模观测器实现系统状态观测,得到观测器误差;利用所设计的观测器对非线性系统进行故障的诊断和分离;采用等效输出注入概念重构了故障信号,使得多变量输入输出非线性系统的故障诊断问题得到了解耦;针对异步电动机系统实现了传感器故障的分离.仿真结果证明了算法的有效性.

Fault detection and isolation of MIMO nonlinear systems based on the cascade observer

This paper considers the application of a block_observable method to the problem of fault detection and isolation.First,a block-observable form of nonlinear systems is introduced.The cascade observer of the system can be obtained by decoupling the system into blocks according to the faults of the sensor and the performer.The observer error can be got through the application of the sliding mode state observer.The observer designed in this paper is applied to the problem of the fault detection and isolation of the nonlinear system.The equivalent output injection concept can explicitly reconstruct fault signals,in which the problem of the fault diagnosis for multi-input and output nonlinear system will be decoupled.The isolation of the sensor fault is obtained in the system of an asynchronous motor,and the result of the simulation validates the algorithm.