Set a reasonable threshold for distance. If the dimers are formed by chemical bonds (including hydrogen bonds), there may also be restraints on the bond angles.
Sometimes (kinetic energy minus potential energy) is used to distinguish stable pairs from accidental collisions at high velocities.
Your interaction potential probably has a minimum. Yiu can use the distance of that minimum + 20%, for example. Or (better!) you can calculate the distance at which the pair potential is higher than the mimum value by an amount corresponding to the avarega kinetic energy.
I am simulating a system by placing two monomer at a distance with some designed force field. And now trying to calculate some property or values that can prove the formation of dimer.
If you have a bonded-interaction force field and the non-bonded interaction is not repulsive and outweighing it, you must have an energy minimum somewhere.
You have got a force field for the interaction of two monomers, i.e., a function that computes the potential energy for given distance (and perhaps orientation variables). This function will run to zero for large distances and to +infinity for short distances. If a stable dimer is formed, this function must have a minimum somewhere.
If you expect a dimer to exist, but do not find a minimum, you are using the wrong force field.
I am getting the minimum energy with respect to number of simulation moves. But this minimum energy may not be predicted initially, only after the rum gets completed.