曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
7.3. PERFORMANCE
tone. This creates modulation sidebands on each side of the meshing tone, with distance to the carrier equal to the shaft rotating speed. As unbalance increases, so does the modulation sideband energy, which is captured by the MOD indicator. Any fretting between the gear surfaces causes an increase in random noise, which increases the noise .oor of the signal power spectrum. This phenomenon is captured by the RMSR indicator. Consequently, the indicators MOD and RMSR are chosen for detecting this fault type.
To detect the presence of faults, simple thresholds are applied to the expected value slope ac.and the scatter level slope ag.w for each indicator. Scales 1 to 8 are chosen both for ac.and ag.w . Thresholds are based on the .uctuation-envelopes for cases 17 to 20. From these cases, the min and max values for each scale of ac.and ag.w for each of the two indicators are retrieved, and used to form the threshold basis. The actual thresholds used for fault detection are the threshold basis multiplied by a factor. I.e. if the threshold factor is set to 120%, an alarm is generated whenever ac.or ag.w is more than 120% above the max value or 120% below the min value experienced in the normal state training set.
Case 2 is illustrated with indicator decomposition, noise level estimate and
slope
shown
in
(Fig.
7.3,
7.4
and
7.5)
for
MOD
and
(Fig.
7.6,
7.7
and
7.8)
for
RMSR.
The
dotted
lines
are
the
threshold
bases
for
each
scale.
Only
scales 4 -6 are plotted, in order to make the .gures more readable.
A fault detection test on all the cases is conducted using threshold factors ranging from 90% to 150% in 10% steps, with results summarized in (Fig. 7.1).
The "Length" column contains the length, in .ight hours, of each data set. The "HUMS" column contains the ground-station detection results for each case, with the ground-station using traditional learned thresholds. Four of the fault cases (1, 3, 12 and 14) were missed by the ground-station diagnosis method, and were discovered by either metal chip-alarms or routine inspections. Case 7 was discovered by the operator manually inspecting the indicators and signals. It can thus not be known if the HUMS would have generated an alarm, had it not been detected manually.
It is not known if, and how many, false alarms were generated by the healthy state datasets. As a global average, a HUMS generates alarms in the magnitude of 4 to 12 per 1000 .ight hours. With the component in question being one of about 80 components monitored on the AS332, it is likely to believe that it would produce a proportional number of false alarms.
7.3. PERFORMANCE
Case State Length HUMS 90 100 110 120 130 140 150
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：OPTIMIZATION OF FAULT DIAGNOSIS IN HELICOPTER HEALTH AND USA(65)
