1. d6 Qf8+
2. Kh7 Qd8
3. Qg4 Kd6
4. Qe6+ Kc7 (K? Qe8)
5. Ne8+ Kb7 (Kb8 Qd6+)
6. Qd6
Ka7
7. Qc7+
Q?
7. d8
Q?+
7. Kg8 Q?+
8. Kf7
Qd8
9. Ke6 K?
10. Qc7
Q?+
9. Ke7 Qg5+
10. Kf6 Qg7+
11. Ke6
Ka8
7. Qc7.

Those are the main variations I can think of that don't involve blunders.

1.d7 you mean. My line would be 1.d7 Qh3+ 2.Nh5! Qxh5 3.Kg8 Qh4 4.Qd1+ winning

@Eugene:

If your first move is 1. d6-d7, Black can play 1. ..., Qf3-h3†and the pawn drops off.
Except if White plays 2. Ng7-h5!! when taking the pawn is unwise because of the knight fork.And if 2. ..., Qh3xh5† then 3. Kh8-g8! and Black has no sensible check.

I looked up in a Nalimov tablebase if this queen ending was won, thereby depriving my own imagination the chance to find a further sparkle (if I could have found it at all).
The ending is indeed won. Black's best continuation is
3. ..., Qh5-h4
4. Qg1-d1†, Kd5-e6
5. d7-d8R!!
The only move to win.
5. ..., Qh4-g5†
6. Kg8-h7
Should White have promoted to a queen, thereby covering the g5 square, he could not have captured the black queen anyway because of stalemate.
6. ..., Qg5-h4†
But here the rook promotion makes a real difference. With a queen on d8, Black could have played 6. ..., Qh5†! with a new stalemate if white captures, or with perpetual check otherwise.
7. Kh7-g6
Quite amazingly, the White King can ultimately escape perpetual check, if he seeks refuge at... b2(!). Interesting to replay this stage. Black's problem is that White can sometimes interpose his queen on g4 and later b3, which would force a queen trade because of the position of Black's king.

@MvE. Ah, you're right. Sorry, my chess engine didn't catch that, guess this Endgame is a bit too deep for Crafty to calculate (as Black) with a timer on (I'm kind of impatient).