[Ipopt] Non-optimal solution

Pratim Vakish pratim_usc at hotmail.com
Sat Nov 29 12:46:37 EST 2008


I am trying to solve, with IPOPT, a continuous non-linear optimization problems which include the following constraints and variables (code in AMPL below):

var ga{i in S} >=0, <=1;
var w{i in S} >= 0;

subject to we1 {i in S}:    w[i] >= 0.0001* ga[i];
subject to we2 {i in S}:    w[i] - ga[i] <= 0;
subject to we3{i in ST}:    (ga[i]-0)*(ga[i]-1) >= 0;

IPOPT terminates (fast) by telling me that it found the optimal solution.
However, the solution found IPOPT is not optimal. It can be verified easily on small problem instances.

What is it due to? How could I solve (slightly modify the formulation, scaling, etc.) this issue? 



Get more done, have more fun, and stay more connected with Windows Mobile®. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.coin-or.org/pipermail/ipopt/attachments/20081129/52a71c92/attachment.html 

More information about the Ipopt mailing list