Pseudo power signatures for aircraft fault detection and identification

In earlier work, we presented proof of concept for the use of signal processing tools for fast fault detection in a model free framework. In this paper, we elaborate on the concept by applying pseudo power signatures to detect faults. We introduce a Singular Value Decomposition (SVD) approach for the computation of pseudo power signatures and discuss some of the advantages of the approach. In particular, we expand the concept of pseudo power signatures to that of signature subspaces. New experimental data from a reliable B747 aircraft model are used to test the proposed schemes.