The main line goes as follows 

Any suggestion? 
So far this is what I (the computer) came up with (with quite a bit of sidelines)so far

no. this is not going to work for you and its quite typical of computer inspired openings. I think 11. Bd6 is an extremely unlikely move. the one other line you looked at, qc7, that line goes absurdly deep, seriously, with very few of the moves actually being forced (same is true with the bd6 line, indeed especially that one, because black has shed a piece, he can be as creative as he wants heh). i think you have reasonable chance of reaching the position at move 11, but then the game may go all sorts of directions. IN my opinion, this central advance looks pretty strained, and i dont trust it. I would at least castle first. Even then,its hard to tell if the handing pawns are good for you or not, blacks pieces certainly dont seem to be poorly placed and i think he could blockade your center and make it weak, very likely.I dont know if you are familiar with more common minority attack ideas, but this would seem like a nice option. (b4-b5 plan) in this position there are too many possible moves, you cant just know all of them, what you need are good working plans and thematic ideas.

Castleing first runs into 

Ideas are much more important than move orders in the queen's gambit since both sides have so many plans available. This isn't the kind of opening you want to learn by rote. Your results will be very bad.

black trades some pieces, sure, white cant get everything in the QGD, this is why its considered a solid opening for black. Better to allow some exchanged imo (and like i said, blacks pices are not very badly placed) than to create some pretty clear weaknesses for black to jump onto.

I know the ideas fairly well (minority attack and advance in the center and sometimes Ne5 followed by f4)
The problem with the h6 line is that it's hard to prevent the exchange of dark-squared bishops and if the bishops are exchanged e3 becomes too weak after f3 leaving White with only the minority attack which also isn't as effective as in the line where the knight is on f3 because the knight is more active there. Also it blocks the retreat square of the bishop if Black ever plays c5-c4, forcing it to go to f5 where it will likely be traded as well.

So the only other option is 0-0-0 with a kingside attack but those lines don't promise much either (along with the fact that it's not so much in tune with my style of play.).

The problem isn't so much that Black trades pieces, it's that by doing so it prevents one very important plan for white after which whitee doesn't have much of a chance for an advantage.
Now I'm considering a move order like 

after black frees his game by forcing some exchanges (a normal qgd thing), i dont think he has quite equalized yet. White has the c file, black the e file. White can create pressure on the c file by advancing the b pawn, while it is much harder for black to advance his f pawn. There are other ideas also, but this is the really critical one, imo.

sorry i posted my last before i saw your replies. This h3 idea you had looks more interesting. But idk, there must be some stuff know about this, how have any games gone so far? im not a 1d4 player nor do i usually play the qgd, i only know the very basics, but have good general positional feel. h3 looks interesting, black still cant quite manage to push his f pawn, i dont think.

maybe you could play your idea with ne2 and f3, but then NOT play e4 for quite some time, instead bring bishop to f2, castle. that could be interesting. I think you need to be ok with playing very sharp that way however, maybe even 0-0-0 in some sitruations.

In the main line White doesn't push until everything is prepared ron e1 or d1 h3 a3 and b4 to prevent c5 sometimes and Kh1 I don't think it'll work with Ne2 though since the inclusion of h3 weakens the dark squares quite a bit.