5.3. INDICATOR CORRECTION
T ).1 T
a.=(pepe pe .(i . μi) (5.5)
Figures
5.1,
5.2
and
5.3
shows
the
left
hand
free
wheel
gear
rms
indicator
from an EC225 in slow cruise as a function of indicated airspeed, lateral pitch, and pitch attitude. Indicated airspeed is the speed of the aircraft relative to the atmosphere, lateral pitch is the cyclic stick position in the lateral (forward) direction, and pitch attitude is the angle of the aircraft in the forward direction. The unit is knots for the .rst parameter, and percentage of max angle for the two others. During cruise, these three parameters are strongly correlated. The data for this example was however acquired in slow cruise. In such condition there is a substantial delay between a change in stick position, subsequent change in aircraft angle, and .nally change in aircraft speed. The three parameters are thus only partially correlated for this dataset. Common for all three parameter is however their correlation with engine torque.
From the .gures, it appears as if there is a strong correlation between all three parameters and the indicator. The green line in each .gure shows a third order polynomial approximation of the relationship between indicator and parameter.
Figure
5.4
contains
the
above
indicator
as
a
function
of
acquisition
index,
with the raw indicator accompanied by a corrected one. Model estimation was done using acquisitions 50 to 110. As can be seen from the .gure, the model remains valid also outside the training period.

The relationship between indicator values and environmental conditions is speci.c to each indicator and aircraft. In the above example, there is a certain correlation between the three parameters. All three parameters are also known to have an impact on torque, which is probably the underlying cause of the indicator .uctuations. Physical models for rotorcraft dynamics and their impact on vibration signatures are however outside the scope of this study.

5.4 Signal Correction
Working directly on the signals, it is only necessary to estimate one model per component, although this argument is countered by the increased com-plexity of this method. An advantage held by direct signal correction is that it is producing corrected raw signals suitable for non-linear indicators and classi.cation
methods
working
directly
in
the
time
or
frequency
domains
[46].
Examples
of
such
are
adaptive
lifting
[41]
and
mathematical
modeling
[14].
　
中國航空網(wǎng) www.k6050.com
航空翻譯 www.aviation.cn
本文鏈接地址：OPTIMIZATION OF FAULT DIAGNOSIS IN HELICOPTER HEALTH AND USA(41)