[AMPL 18381] MINLP Solvers cant solve my problem

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

[AMPL 18381] MINLP Solvers cant solve my problem

Murilo Souza
Hello, i've tried with both Baron and Couenne to solve my MINLP problem, and both of them failed. It's a problem of optimal allocation (through a binary variable, a) of certain equipment. I'm 100% sure my problem isn't infeasible, because i've tested it without the optimization part (by trial and error) and i can find several solutions to it.

I've attached the .mod, .dat and a .run files, if anyone can see what might be the issue here, on why they are failing to get a viable solution.

For example, in my "manual" code, without the optimization, if i enforce that one, and only one equipment, should be allocated in the 10th position, i.e:

a[10] = 1;

and enforce a value to hs, like this:

hs[10] = 5 (i.e hsmin = hsmax = 5)

I should be getting a viable solution.

What could be the issue here? Maybe while running the Branch and Bound tree, the solvers are removing some parts of the search espace?

Thanks in advance!

--
You received this message because you are subscribed to the Google Groups "AMPL Modeling Language" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
To post to this group, send email to [hidden email].
Visit this group at https://groups.google.com/group/ampl.
For more options, visit https://groups.google.com/d/optout.

codigo_17_barras.mod (5K) Download Attachment
dados_codigo_17_barras.dat (35K) Download Attachment
script_codigo.run (228 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

[AMPL 18387] Re: MINLP Solvers cant solve my problem

pietro belotti
Can you send the .nl file for the problem? It can be generated after the "data" command in script_codigo.run with the line

write gprob;

which results in a prob.nl file.

Thanks,
Pietro

On Thursday, March 14, 2019 at 5:24:43 PM UTC, Murilo Souza wrote:
Hello, i've tried with both Baron and Couenne to solve my MINLP problem, and both of them failed. It's a problem of optimal allocation (through a binary variable, a) of certain equipment. I'm 100% sure my problem isn't infeasible, because i've tested it without the optimization part (by trial and error) and i can find several solutions to it.

I've attached the .mod, .dat and a .run files, if anyone can see what might be the issue here, on why they are failing to get a viable solution.

For example, in my "manual" code, without the optimization, if i enforce that one, and only one equipment, should be allocated in the 10th position, i.e:

a[10] = 1;

and enforce a value to hs, like this:

hs[10] = 5 (i.e hsmin = hsmax = 5)

I should be getting a viable solution.

What could be the issue here? Maybe while running the Branch and Bound tree, the solvers are removing some parts of the search espace?

Thanks in advance!

--
You received this message because you are subscribed to the Google Groups "AMPL Modeling Language" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
To post to this group, send email to [hidden email].
Visit this group at https://groups.google.com/group/ampl.
For more options, visit https://groups.google.com/d/optout.
Reply | Threaded
Open this post in threaded view
|

[AMPL 18388] Re: MINLP Solvers cant solve my problem

Murilo Souza
Yes, here it is.

One thing to note:

- I'm testing with w1 = w2 = 0, wich means, i'm only minimizing the summation of the allocation of equipments, through the summation of the binary variables a.

I've tried with different values of barramin, barramax, hsmin and hsmax.

Most of those tries, it says my problem is infeasible. But sometimes, it kept looking for a solution, but saying something like:

Cbc0010I After 100 nodes, 26 on tree, 1e+050 best solution, best possible 2 (510.87 seconds)

Wich doesn't make sense to me, since i know theres a better solution than that.

Em sexta-feira, 15 de março de 2019 09:04:35 UTC-3, Pietro Belotti escreveu:
Can you send the .nl file for the problem? It can be generated after the "data" command in script_codigo.run with the line

write gprob;

which results in a <a href="http://prob.nl" target="_blank" rel="nofollow" onmousedown="this.href=&#39;http://www.google.com/url?q\x3dhttp%3A%2F%2Fprob.nl\x26sa\x3dD\x26sntz\x3d1\x26usg\x3dAFQjCNGU992nq3CK0KAqcDmj-24x3DUQ3g&#39;;return true;" onclick="this.href=&#39;http://www.google.com/url?q\x3dhttp%3A%2F%2Fprob.nl\x26sa\x3dD\x26sntz\x3d1\x26usg\x3dAFQjCNGU992nq3CK0KAqcDmj-24x3DUQ3g&#39;;return true;">prob.nl file.

Thanks,
Pietro

On Thursday, March 14, 2019 at 5:24:43 PM UTC, Murilo Souza wrote:
Hello, i've tried with both Baron and Couenne to solve my MINLP problem, and both of them failed. It's a problem of optimal allocation (through a binary variable, a) of certain equipment. I'm 100% sure my problem isn't infeasible, because i've tested it without the optimization part (by trial and error) and i can find several solutions to it.

I've attached the .mod, .dat and a .run files, if anyone can see what might be the issue here, on why they are failing to get a viable solution.

For example, in my "manual" code, without the optimization, if i enforce that one, and only one equipment, should be allocated in the 10th position, i.e:

a[10] = 1;

and enforce a value to hs, like this:

hs[10] = 5 (i.e hsmin = hsmax = 5)

I should be getting a viable solution.

What could be the issue here? Maybe while running the Branch and Bound tree, the solvers are removing some parts of the search espace?

Thanks in advance!

--
You received this message because you are subscribed to the Google Groups "AMPL Modeling Language" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
To post to this group, send email to [hidden email].
Visit this group at https://groups.google.com/group/ampl.
For more options, visit https://groups.google.com/d/optout.

prob.nl (72K) Download Attachment