Figure 1 shows the line has been present for a few days and then disappears on Apr 18. I have checked back in data and the line was also present in the last week of March.
Figure 2 is a zoom of the spectrogram for Apr 18 showing the transition is around 8-9 UTC.
Figure 3 shows the spectrum before and after the disappearance. The line disappears from h(t) and DARM.
Looking in the logbook the only action is discharging the 50Hz feed-forward filters , but it is hard to know if this is the cause, as the time of the 50Hz FF worked is not written down in the logbook. It would be interesting to investigate if this, or any other action caused the 186Hz line to disappear.
The effect of the 50 Hz feed forward is visible in the probe Sc_WI_MIR_Z_CORR (actuation on the mirror). Any bad side effect of 50HzFF should be correlated to something in that probe.
Concerning the period of malfunctioning and the consequent reset, it is well traced by the amplitude in DARM or hrec of the 50 Hz itself. According to the spectrograms, the 186Hz disappeared well before the feed forward restoring.