Papy wrote:I undersrtand and I think that it was the idea which my théory need

I think gsf has emphasize many many times that this method could only be used to prove 2 puzzles being different, but not isomorphism (i.e. 2 grids with identical #pencilmark list could still be different)... I hope someone would show us such an example but I doubt gsf would ever post here again...

You need extra methods to prove if 2 grids are isomorphic to each other... You should research on this link for the exact algorithms from the experts:

http://www.setbb.com/sudoku/viewtopic.php?t=795