Question: since the equation number is mainly a tag for referencing, I usually omit the equation number whenever an equation is not referred throughout the paper. Is this in violation of any standards, in particular those of IEEE transactions?
since the equation number is mainly a tag for referencing,
In fact, a reviewer or someone else writing another paper might need to refer to any of your equations, even those you don't refer to. Thus, better number all the equations.
The IEEE Editorial Style Manual specifies (p. 16):
Consecutive Numbering: Equations within a paper are numbered consecutively from the beginning of the paper to the end.
Note also that the typesetter would probably put all the numbers anyway.