[Ipopt] optimality of solution

[Manuel Andres Ramos Castillo]

If the model is nonconvex, the only way to assure that is trying to solve
the optimization problem with different initial points (better if these are
logical with the constraints). On the other hand, and more if you have a
nonconvex model, you are lucky if you can find a local minimum/maximum. 

 

De: ipopt-bounces at list.coin-or.org [mailto:ipopt-bounces at list.coin-or.org]
En nombre de Hossein Haghighat
Enviado el: martes, 18 de octubre de 2011 07:19 a.m.
Para: ipopt at list.coin-or.org; ipopt-request at list.coin-or.org
Asunto: [Ipopt] optimality of solution

 

Hello,

I have solved a nonconvex EPEC (equilibrium model with equilibrium
constraints) model with Ipopt, using two modeling approaches involving:

1- an NLP approach, where the problem is reformulated as a set of nonlinear
equations,  

2- a diagonalization approach where the problem is reformulated as a set of
MPECs (mathematical problem with equilibrium constraints)   

 

I get the same solution from these different approaches which supports the
optimality of the solution. 

Is there any other way to ensure that the solution is indeed optimal not a
saddle point or a local maximum? (e.g by using IPOPT options, or initial
point manipulations)

 

thank you in advance,

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/ipopt/attachments/20111018/45d4f45e/attachment.html>