Dear Simin,
due to the fact that the fluctuations are proportional to the square root of the number of
particles your pressure as a function of time seems to me very reasonable. I hence,
would only suggest to substantially reduce the timecon to a much lower value, e.g.
50 - 100fs. Also, please keep in mind that the region in the thermostat section is by
default global, which to the best of my knowledge, only couples to 3 degrees of freedom
of the whole system! So better use massive to equilibrate your system.
Best,
Thomas