Abstract:
A technique for the identification of aircraft stability and control parameters from
flight test recordings made in either calm or turbulent air is presented. The maximum
likelihood output error method is used with a steady-state Kalman filter incorporated
to account for atmospheric turbulence. A modified Newton-Raphson search technique,
enhanced by a line search, is employed for parameter identification. Separate
algorithms are developed for estimating the biases and noise levels in the observations.
Some areas of practical problems in the application of such methods are stressed. A
computer program for the identification of longitudinal stability derivatives is
described and the aircraft instrumentation required is exemplified by that in Gnat
XPSOS. The wind tunnel calibration of the flow-direction sensing nose probe
assembly of Conrad yawmeters on this aircraft is detailed. The problems of handling
the flight observations recorded are covered and some of the troubles experienced
with the instruments are noted.
The performance of the identification technique is investigated. The data required,
the choices open to the analyst and the statistical information produced being highlighted.
The identification of the process noise level, in this instance the turbulence
intensity, is addressed and it is found that the value specified for this level can
influence the other parameters.
The longitudinal stability and control derivatives obtained for Gnat XPSOS are
presented.
