Detecting and identifying faults at sensors for state values - using association matrix giving optimal relationship between characteristic vectors and associated classification vectors

摘要

The condition values are coupled to the measurement values directly delivered by the sensors by a measurement equation m=Hx+ epsilon, where m is a vector of the measured values, x a vector of the state values and H the measurement matrix. The rank of m is greater than that of x. That is a general description os a similar or dissimilar redundant measurement. Characteristic vectors, vi, are set as gap vectors of a characteristic matrix. A parity vector, v, is taken as an element of the parity space that is an orthogonal complement of the signal space. This is defined by a linear combination of the characteristic vectors, vi, and the associated elements, mi, of the measurement vector m. USE/ADVANTAGE - High reliability condition meaurements as necessary in aircraft, nuclear power stations or chemical processes, the sensors involved being gyros, accelerometers, Doppler radar speed measurer, thermo-element for temp-level. Can detect deterioration of sensor signal and categories as "hard, mid-value of soft failure".