基于多指标约束的鲁棒故障检测方法

Robust fault detection based on multi-index restriction

针对存在模型不确定性和测量噪声的线性动态系统,研究了其鲁棒故障检测问题。通过H∞性能指标和无故障时残差的稳态方差约束反映对干扰影响的抑制,采用H-性能指标刻画对故障的灵敏度,将故障检测滤波器的设计描述为一个多指标约束问题,并利用线性矩阵不等式方法进行了求解。最后,在一种应用捷联惯性导航设备的车载卫星天线稳定跟踪系统中进行了仿真研究,结果表明,这种方法能够有效提高系统的故障检测精度。

In view of the model uncertainty and measurement noise in a linear system, the robust fault detection is studied. The influence of disturbance on the residual is described by applying H_ norm and the fault-free residual stable variance constraints, and the sensitivity to fault is described by the H_ norm, therefore the design of fault detection filter can be described as an multi-objective restriction one. The solution is abstained by solving （LMI）. The simulation on stabilized antenna system carried by vehicle is made for mobile satellite using SINS. The result shows that the way can effectively improve the precision of fault detection.