Fiktor, many journals explicitly indicate option (1). Some also ask to use single/double numeration, the list of references in alphabetical/appearance order, and many other things. Therefore,
Strategy (4): check with the journal you are planning to publish your contribution.
As for (2) and (3), are you happy if somebody else cites your ms in the form [1, Eq. (7.15)]? Wouldn't it be better if there is no number and he needs to reproduce your formula with link to your original source?
Strategy (2) is standard for physical journals, so if you are more on the math physics side, choose (2) or (3) without doubts.