一种基于LMI的鲁棒故障诊断滤波器设计

研究了一类含有扰动的线性系统鲁棒故障诊断滤波器设计问题. 文中引入一种能同时体现残差对扰动信号鲁棒性和对故障信号灵敏性的性能指标, 利用H_∞理论把求解滤波器的问题转化为H_∞优化设计问题. 应用线性矩阵不等式(LMI)技术, 对此性能指标进行优化, 给出并证明了该设计问题解的存在性条件和滤波器增益阵的求解方法, 最后, 通过仿真实例验证了方法的有效性.

The design of a robust fault-detection filter: the LMI approach

The design of a robust fault-detection filter(RFDF) for a class of linear systems with disturbances is studied. A performance index which includes both robustness against disturbances and sensitivity against faults is introduced. By using some results of H-infinity optimization, the RFDF design problem is formulated as an H-infinity optimization problem. The performance index is also optimized via the linear matrix inequality(LMI) technique. Moreover, the existence conditions for solutions are derived and the gain matrix of the above filter is determined. Finally, an example is given to demonstrate the effectiveness of the proposed approach.