Diversity of types of information faults in navigation complexes makes the development of algorithms for fault detection and isolation to be realized in the on-board computer complicated. Moreover, a complete list of possible types of faults may be unknown. Two algorithms are suggested to overcome these difficulties. The first algorithm presumes a set of types of faults to be known. It employs an augmented Kalmanfilter (AKF) that estimates a group of possible faults and allows the amount of computations to be reduced. It has been shown that the results of the AKF application allow derivation of a posteriori probabilities and estimates of particular faults as well as estimates of additional errors in navigation parameters caused by faults. These estimates can be used to recover the faulty system. The second algorithm does not require knowledge about types of faults and is intended to detect the system of the complex that has errors, anomalous in level and behaviour, and to compensate for these errors. The efficiency of the algorithms developed is supported by the results of simulation and bench tests of the navigation complex including more than one inertial system.
Economical algorithms for detection and isolation of information faults in navigation complexes
Ein effektiver Algorithmus zur Aufdeckung und Isolierung von Informationsfehlern in einer komplexen Navigationsumgebung
2007
8 Seiten, 6 Bilder, 9 Quellen
Aufsatz (Konferenz)
Englisch
Prediction Algorithms in Economical Processes
British Library Conference Proceedings | 1999
|Economical power supply with galvanic isolation
Tema Archiv | 1976
|Robust Detection and Isolation of Process Faults Using Neural Networks
British Library Online Contents | 1997
|Multiples Faults Detection and Isolation in Sensors of Dynamic Systems
SAE Technical Papers | 2005
|