I want to inform you that this is a theoretical endgame and the double pawns are not such a big weakness as you might think. In GM play most of the times ends up in a draw.
Now an interesting way to play since in comparison with the Philidor endgame here you don't have omit Nf3 is to develop your knight via h3. Try the maneuver Nh3-Nf2 and in coordination with Bg5 you might find something to put pressure. Don't give up the Bishop on f6, retreat and put pressure on e5 pawn. That will keep Black pieces more tight but an engine will have no problem to defend. At least you will not have to worry how to keep your pawns intact.
I'm being repeatedly humbled by Computer Level 10 in this position from the Pirc Defense, where I'm playing White. I'd swear White has a significant advantage here, and with little danger, since Black has doubled isolated e-pawns, Black can't castle, Black has lost at least one tempo from my "queen slap," and there are no queens on the board. But evidently I don't know how to exploit those doubled pawns, or where I should put my pieces, or in general what my plan should be, so then the computer finds a way to win. This is another of those endgames where there are many pawns on the board and often I just don't know how to handle such positions.