Title :
Sensor failure estimators for detection filters
Author_Institution :
Rolls Royce plc., Derby, UK
fDate :
10/1/1992 12:00:00 AM
Abstract :
The processing of (detection filter residuals) in order to detect and identify sensor failures is considered. Simple projection techniques are shown to be inefficient, so a single sensor failure-tracking estimator is derived. This estimation filter uses the planar nature of the detection filter residuals resulting from a sensor failure to improve the performance of both the detection and identification of failures. The performance improvement is illustrated using data from a simplified model of a turbofan engine
Keywords :
failure analysis; filtering and prediction theory; signal detection; failure detection; failure identification; sensor failure-tracking estimator; turbofan engine; Actuators; Engines; Equations; Failure analysis; Fault detection; Filters; Sensor systems; State estimation; Transmission line matrix methods; Vectors;
Journal_Title :
Automatic Control, IEEE Transactions on