《Fault Detection for Linear Discrete-Time Invariant Systems with Decoupling and Optimization》.pdf

Joint 48th IEEE Conference on Decision and Control and WeAIn6.1 28th Chinese Control Conference Shanghai, P.R. China, December 16-18, 2009 Fault Detection for Linear Discrete-Time Invariant Systems with Decoupling and Optimization Xiaobo Li and Kemin Zhou Abstract— This paper discusses the robust fault detection has also been derived in [3] for some different objective design for the criteria such as H /H , H /H and H /H − ∞ 2 ∞ ∞ ∞ that minimizes the ratio of disturbance sensitivity and fault in which G , the transfer function from disturbance to the d sensitivity. Furthermore, [11] and [12] have extended these measurement output, is a tall transfer matrix and Dd is not full column rank. It is shown that the faults in a subspace can be results to time-varying systems based on some time domain made arbitrarily sensitive, while the faults in the complemen- formulations. However, in those papers [9], [10] [3], [11] tary subspace have bounded sensitivities that are maximized by and [12], it is assumed that Gd , the transfer function from our filter. We also discuss the decoupling and non-decoupling disturbance to the measurement output, is a square or wide conditions, and their relations with image spaces of Gd and transfer matrix and Dd matrix is of full row rank, which is Gf (the transfer function from disturbance to the measurement conservative in some sense. output and the transfer function from f