Estimation of Periodically Occurring Faults for Dynamic Systems With Timing Misalignment Between Faults and Observations

In this article, the challenging task of estimating periodically occurring faults (POFs) in sampled-data systems is investigated. One major obstacle faced in this study is the timing misalignments between fault occurrences and observations, which is a common issue in practice. A novel approach to modeling the influence of the POFs is proposed, which utilizes an important auxiliary variable indicating the remaining time to the next fault occurrence. A fault estimation approach is designed based on the particle filtering to address the difficulty caused by the complexity of the proposed model. Moreover, a performance evaluation method is also applied to improve the precision of the estimation. Finally, an experimental study is conducted based on a kind of autonomous underwater vehicle, which illustrates the advantages of the proposed method.