Optimization of Non-Quadratic Functions

Dear All,
I am trying to implement some portfolio optimization techinique going beyond mean/variance optimization (e.g. including higher order moments in the quantity to maximize or using non-quadratic utilities). Furthermore, I am also interested in applying nonlinear constrains to my portfolio weights.
For all these tasks, I need something different from the solve.QP routine I have been using happily so far.
Is there any specifically designed routine for this kind of problem or should I revert to some general-purpose optimization package?
Any suggestion is welcome.

Best Regards

Lorenzo