Hi Naeiji,
Your formula is of an implicit function, which Origin currently doesn't support for fitting. For now, possible work-arounds (practically, but NOT theoretically to give you the sense of parameter values) may be the following either way:
1) You can rewrite your function as X as dependent, and Y as independent. Thus, x = (y/(1-y)*exp(2*a*y))^v/b
(Of course, the parameters would minimize the Chi^2 on X, instead of Y.)
2) Define a function:
Z=y/(1-y)*exp(2*a*y)-(b*x)^(1/v)
and add an extra dataset for Z as the dependent variable which have all zeros(0) corresponding your (x, y) values (as independent).
Hope this helps.
--Hideo Fujii
OriginLab