To my knowledge there's no engine that's ready-made to do sims 2-like gameplay.
And let's be honest here: Creating a sims 2-like game engine (on top of normal game engine) is too ambitious for a one-man project. And that's the programming/testing side of it alone, you'd also need to model and animate characters/objects/locations/etc. Which would make it even more overambitious.
Now there is one suggestion I have, if you want to stick to this kind of gameplay: create it as a mod for The Sims 2/3 (as far as I know TS3 supports C# scripting, not sure what TS2 offers). That way you could focus entirely on your story (although there's probably some challenges like getting the right people in the right place at the right time, if it is actually possible to do stuff like that). Depending on what your ambitions are you may need to do some animation, though there is some sexual stuff available in mods.
I think that's your best bet if you want that kind of gameplay.
The alternative is to lower your ambitions, especially if this is your first project: start smaller! Consider that shark takes at least 4 months to create a game, and that's with all the experience he already has. On top of that, the scenes in shark's games are fairly static. What you're proposing is a lot more dynamic, and consequently it's a lot more work and a lot harder to test.
To make anything in 3D look good is a tough job (you might be able to make some headway with something like exporting DAZ models to Unity, like this:
http://www.youtube.com/watch?feature=pl ... LhTbSJecyI , though I have no idea what the licensing situation is like for stuff like that).
All in all I think one-man projects are most likely to succeed if it's kept small. Even creating a simple HTML dating game with static graphics from DAZ (for example) will take a ridiculous amount of time to create...