in npt process can not keep constant pressure

Dear all,
I need urgent help regarding fixinf NPTsimulation,
I actually running a methane hydrate cage simulation but
i can not keep constant pressure.
please help me.

thanks in advance,
snehanshu
research scholar , IIT Kharagpur
command/control file
is paste below:

=====================================

512 + ONE METHANE 2X2X2 BOX

=====================================

units real
boundary p p p

atom_style full
bond_style harmonic
angle_style harmonic
pair_style lj/cut 10.0
pair_modify mix arithmetic
#kspace_style pppm/tip4p 1.0e-6

read_data 51262MFINAL2.txt

pair_coeff 1 1 0.13389 3.166
pair_coeff 1 2 0.01716 2.61
pair_coeff 3 3 0.42566 3.35
pair_coeff 1 3 0.11767 3.258
bond_coeff 1 15.3299 0.9572
bond_coeff 2 4.97 0.9572
angle_coeff 1 0.0 109.52
angle_coeff 2 0.0 109.52

neighbor 2.0 bin
neigh_modify delay 0 every 1 check yes page 10000 one 1000

thermo 10
thermo_modify flush yes
minimize 1.0e-5 1.0e-7 1000 10000

dump 1 all atom 100 dump.lammpstrj
dump_modify 1 flush yes
log run1.txt

#fix 1 all shake 1.0e-6 200 0 b 1 a 1

=====================================================

equilibration: NVT

=====================================================

fix 2 all nvt temp 200.0 200.0 1.0

timestep 0.1
velocity all create 300.0 34239
run 5000

=====================================================

equilibration: NPT

=====================================================

unfix 2
fix 2 all npt temp 200.0 200.0 1.0 iso 15.0 15.0 0.1
timestep 0.01
run 5000

=====================================================

dump atom positions for visualization, e.g. using VMD

=====================================================

#timestep 1
#run 5000

In small systems is normal the pressure fluctuate a lot. If you calculate the mean value of the pressure you will see that the mean value is in agreement with the desired value.

Rodrigo Freitas