Abstract:

Mean value test (MVT) is proposed to detect and isolate the failure. Non linear filtering is also proposed to estimate the combinations of sensor errors affecting the parity vector, the estimates are used to form a compensated parity vector. The compensated parity vector is used to make FDI decision for improving FDI performance. To illustrate its application, the MVT approach is used to detect and isolate the failure for an optimal mounting configuration of five single degree of freedom sensors. Some results are obtained by comparing MVT method with GLT method. Simulation results prove that MVT method can detect and isolate the failure effectively when two sensors fail simultaneously.

Key words: failure detection and isolation, mean value test, non linear filter, analytical redundancy