how to well equilibrate a system after heating

Dear all:
As we can see at many places, a system ,after heated or cooled to a certain
temperature, should be at equilibrium prior to post-studies. For lammps, 'fix npt'
and 'fix nvt' can be used to thermostate a system, but how can we ensure a
system is well equilibrated? Is it feasible to use 'fix nve' or use large enough
relaxing timesteps for 'fix npt/nvt' to well equilibrate a system?
Any suggestion will be deeply appreciated.

L.Yang

Monitor the temperature and other thermodynamic quantities
during a long NVE run. If they are stable, the system is stable.

Steve

Dear Steve:
  Thanks for your response.

Here are parts of my simulation using NVE to equilibrate a system with boundary conditions 'psp'.

As we can see, after 100,000 timesteps equilibrating, the temperature fluctuates aroud 1390k, and ly seems to be
equilibrated. I am not sure whether this system is well equilibrated or not.
Steve, what's your opinion about this ?

It appears to be - but you need to learn about MD and
statistical mechanics yourself. The mail list is not
a guide as to how to perform MD calculations.

Steve