sci.op-research

thampw@hotmail.com wrote:
> Hi all,
>
> I'm a bit stumped on why Excel solver can't find a feasible solution
> for my optimization problem. I don't think there are too many
> variables and it's an extension of a smaller problem that successfully
> solved.
>
> If you have a bit of time and don't mind looking through my Excel
> spreadsheet, I can explain what I'm trying to solve. BTW, it's a work
> problem.
>
> If anyone is interested to have a look, pls email me at thampw at
> hotmail
>
> Thanks.

First off, do you know that the model is feasible? Can you write down
an obvious feasible solution?

If so, and particularly if you have equality constraints, it might be a
tolerance issue. I don't have Excel on the machine I'm using right now,
so I can't give you the exact parameter names, but if you click into the
settings part of the Solver dialog, there are a few tolerance
parameters. Try making those a bit more liberal and see if Solver finds
a feasible solution.

If you do know a feasible solution and loosening Solver's parameters
doesn't do the trick, you can send the spreadsheet (including your known
feasible solution) to me at the address on this message, and I'll take a
look when time permits.

/Paul