The residue identification algorithm and constrained least squares applied to space shuttle main engine data

Abstract

The residue identification algorithm (RID) is applied to flight data taken from the space shuttle main engine (SSME). Given spectral estimates of the SSME test data and the poles of the system, RID calculates a model of the spectral quantities using a least-squares technique. Application of RID to an SSME spectral estimate is shown. RID is also modified to use constrained least squares (CLS) instead of least squares. This technique is used to obtain models which more tightly fit the data within a user specified frequency range. A contrived example using RID with CLS is included to illustrate the advantages of using this method.