基于卡尔曼滤波器组的多重故障诊断方法研究
Multiple fault detection and isolation based on Kalman filters

针对缺乏有效的用于处理多重(两重及以上)加性故障隔离问题的诊断方法的现状, 本文提出了一种新的基 于卡尔曼滤波器组的控制系统多重故障的检测与隔离算法. 通过构造多个结构不同的卡尔曼滤波器并设计相应的 残差, 使得每个残差仅对执行机构或传感器某个故障敏感而对其余故障不敏感, 最终实现多重故障检测与隔离. 除 此之外, 通过理论推导以及仿真分析, 证明了所提出的故障检测与隔离算法的优越性.
Due to the lack of efficient approaches to locate multiple (two or more) faults, a new method based on a bank of Kalman filters to detect and isolate faults in sensors and actuators is considered in this paper. In the presented approach, Kalman filters are constructed corresponding to all the possible faults of sensors and actuators, and a set of structured residuals is given, each of which is sensitive to a fault and robust for the remaining faults in sensors or actuators. The theoretical analysis of the proposed approach is given. In addition, a simulation example is employed to show the advantage of the proposed approach in multiple fault detection and isolation.