Abstract
This paper introduces a novel online robust fault detection (RFD) framework for continuous-time switched linear systems with unknown disturbances and noises, a method that is set to pique the interest of researchers and academics in the field. First, a dynamic event-triggered mechanism finds out the current mode-dependent event. It updates the active mode based on the sampled information of the unknown synchronous switching signal at each sampling instant. Then, new stability criteria is formulated from both the average dwell time (ADT) and the mode-dependent average dwell time (MDADT) concepts reducing the conservativeness of stability results. This step allows the online switched unknown input observer to estimate the system states immediately. Third, by using linear matrix inequalities (LMIs), the RFD algorithm is designed to identify different unknown faults affecting each current mode in real time. It uses online residual indicators with adaptive thresholds that are robust to deterministic input disturbances. Furthermore, it reduces the impact of stochastic disturbances by reaching H∞ performance with an online switched level. At last, a case study example illustrate that the new schemes have larger feasible regions and more valid for the design of switching signals than the existing results.
Keywords
Get full access to this article
View all access options for this article.
