Tobias Achterberg
unread,Jul 11, 2016, 9:43:11 AM7/11/16Sign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Sign in to report message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to gur...@googlegroups.com
For n=1 you could just minimize or maximize the square of your objective
function, which yields a linear function that is supported by Gurobi. For n >=
2, it is no longer possible to eliminate the square root functions that easily.
Since Gurobi can only handle linear, (convex) quadratic, and piece-wise linear
objective functions, you need to cast your model into such form. One possibility
would be to create a piece-wise linear approximation of your objective function
that is sufficiently close to your true objective function within the
interesting regions of the variables' domains.
Regards,
Tobias