|Abstract:||In this work, we present a fault detection, identification, exclusion, and readmission method in a multi-sensor navigation architecture. The example multi-sensor navigation architecture used in this work comprises of different sensors coupled with an inertial measurement unit (IMU) in individual Kalman filters (KF), where each filter separately provides its navigation solution. These decentralized individual KF solutions are later combined to provide the resultant navigation solution using a weighted fusion. In contrast to a centralized KF where faulty sensor measurements corrupt the navigation solution due to KF memory, the decentralized structure allows fault(s) to be removed from the resultant navigation solution instantaneously after detection and exclusion of the faulty filter(s). Solution separation among subsets of filters is used to detect faults and then identify the faulty sensors or filters, after which they are excluded from the resultant navigation solution. We also provide protection levels against faulty filters, which are identified and excluded. In order to readmit the declared faulty filters, we utilize solution separation techniques again to design a readmission test. This readmission test also allows us to reset a faulty filter in case the fault caused the filter navigation solution to be corrupted. We provide integrity risk and protection level formulas that account for missed detection, and wrong readmission. We validate the developed algorithm for detection, identification, exclusion, and readmission of faulty sensor using real sensor data set and simulated fault scenarios.|
Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022)
September 19 - 23, 2022
Hyatt Regency Denver
|Pages:||1665 - 1675|
|Cite this article:||
Kujur, Birendra, Khanafseh, Samer, Pervan, Boris, Soloviev, Andrey, "Multi-sensor Fault Detection, Exclusion and Readmission," Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022), Denver, Colorado, September 2022, pp. 1665-1675.
ION Members/Non-Members: 1 Download Credit