Co-Designed Multi-Objective Fault Detection Filtering for Networked Switched Positive Systems

초록

This study addresses the multi-objective L-/L-1 fault detection filtering design for switched positive systems (SPSs) subject to a mode-dependent minimum dwell time (MDMDT) constraint utilizing a dynamic event-triggered strategy (DETS). First, a novel linear copositive dynamic event-triggered condition is proposed for systems with positivity constraints. Next, sufficient and necessary conditions are derived to ensure that the designed fault detection filter provides a positive asymptotic estimation with an L-/L-1 performance index. To address the MDMDT constraint, a novel discretized linear copositive Lyapunov function (DLCLF) approach is developed to establish sufficient conditions for ensuring the corresponding augmented system is exponentially stable. Furthermore, by employing the matrix decomposition technology to design filter matrices, an effective mode-dependent piecewise design scheme is proposed for the fault detection filtering of SPSs in a linear programming (LP) form. The developed LP rather than linear matrix inequality (LMI) based scheme features a simpler structure, lower computational demands, and less conservativeness. Finally, three examples are exhibited to validate the achieved results. Note to Practitioners-Faults frequently occur in practical engineering systems, making fault detection essential for ensuring safe and reliable operation. In light of communication constraints commonly present in real-world applications, this paper presents a novel dynamic event-triggered fault detection scheme, in which data packets are transmitted only when necessary. This strategy significantly reduces communication and computational overhead while preserving high detection sensitivity and rapid fault response. Compared with traditional time-triggered and event-triggered approaches, the dynamic event-triggered scheme allows for more efficient monitoring by avoiding unnecessary data transmission during normal operation. Furthermore, the multi-objective fault detection framework is developed to achieve a balanced trade-off among detection accuracy, system stability, and resource efficiency. The proposed method is particularly suitable for embedded monitoring and control systems operating under limited bandwidth and processing capabilities.

키워드