[Ipopt] Inequality Constraint issue

[Stefan Vigerske]

Hi,

if you look at the solution for the model with =e=, maybe it gives you 
some hint on how to set the starting point for Ipopt with the =l= model, 
so Ipopt has higher chances to converge.

Stefan

On 06/19/2013 06:02 PM, reischl maxi wrote:
> Hello all,
>
> i have a problem in a Gams model, using mums as solver.
> If i set the inequality constraint to =l= (lesser or equal to) the solver doesnt find the optimal solution.
> If set to =e= (equal to) the optimal solution is found.
>
> I suspect i have a nonlinear problem because im using PDF and Cumulated PDF of a normal distribution in the objective funtion.
>
> I tried setting
> bound_relax_factor to 0 as proposed in the tipps and tricks list.. it sadly did not help.
> I can confidentially send the gams model if someone wants to take a look at it.
>
> Im grateful for any help.
> Greetings
> Max
>
>
>
> _______________________________________________
> Ipopt mailing list
> Ipopt at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/ipopt
>