Tradeoff design of fault detection systems with high‐order diagnostic observer

Abstract This article is dedicated to the tradeoff design of fault detection with a high‐order diagnostic observer for linear discrete‐time networked systems. Initially, an event‐triggered mechanism with interval limitation is introduced to reduce unnecessary communication transmissions while ensuring timely communication data updates. Subsequently, a high‐order diagnostic observer constrained by Luenberger conditions is constructed to generate residual. Then, using the augmentation method, the residual robustness and sensitivity are comprehensively analyzed to establish a set of performance indicator inequalities, and a performance tradeoff index is proposed as an evaluation metric for observer design. Next, the Schur complement and linear matrix inequality techniques are utilized to develop an algorithm for tradeoff high‐order diagnostic observer that can be robust against the external disturbance and the system input while remaining sensitive to the fault. In addition, a time‐varying threshold is established based on robustness analysis to enhance fault detection. Finally, a simulation example is provided to illustrate the effectiveness of the proposed method.