Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: FAULT DIAGNOSIS OF THE PARALLEL KALMAN FILTER Chingiz Hajiyev 1 1 Istanbul Technical University, Faculty of Aeronautics & Astronautics - Istanbul, Maslak, 34469, TURKEY,[email protected] Abstract - An approach to the test of the parallel Kalman filter is proposed. It is based on the introduced statistic of mathematical expectation of the spectral norm of a normalized innovation matrix. The approach allows for simultaneous test of the mathematical expectation and the variance of innovation sequence in real time and does not require a priori data on values of the change in its statistical characteristics under faults. Using this approach we have developed algorithms for test and diagnosis of the parallel Kalman filter. They consider the technique of multichannel filtering implementation and the number of estimation channels used. Keywords:Fault detection, Fault diagnosis, Kalman filters, Parallel algorithms, Sensor failures. 1. INTRODUCTION The necessity for operative test of the Kalman filter arises in many problems concerned with navigation and control [1, 2]. Some algorithmic methods have been developed for this purpose. The algorithmic techniques for testing the Kalman filter are reviewed in [3] from which it follows that by this way one can ensure not only failure localization and detection but also estimate correction. At present there are number of the algorithmic methods for the testing of Kalman filter [1-7], which is used for fault detection in Kalman filter different diagnostic signs. In spite of the great variety of the algorithmic methods for testing of Kalman filter, till to the present, questions of monitoring and diagnostics of its multichannel modification are not investigated. In many applications it is possible to receive information on the state vector of a dynamic system from several sources simultaneously (as an integrated navigation system). Integrated navigation systems are still used in various applications successfully. In the aerospace and navy navigation systems, GPS, DGPS, GLONASS and INS systems are integrated in different combinations via Kalman filtering [5, 8- 15]. Federated [16-18] or parallel [19-20] Kalman filters are satisfactorily used to integrate different navigation systems. These kinds of filters are known as multichannel Kalman filters. Algorithms have been developed for multichannel estimation of the system parameters and state, which use for estimation a mathematical model of a dynamic system, as well as measurements of several measurement channels (multichannel Kalman filters). In multichannel Kalman filters, simultaneous processing of the available data allows one to improve the estimation accuracy of the state vector and the reliability of data processing. Application of these algorithmic techniques to testing the multichannel Kalman filters is concerned with a considerable increase of the required amount of computation because each estimation channel requires its own "failure detector". Taking into account that the multichannel estimation procedure...
View Full Document

This note was uploaded on 02/18/2010 for the course ITK ETF113L07 taught by Professor Popovskiborislav during the Spring '10 term at Pacific.

Page1 / 6


This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online