Bigger Problem - Less Time??

[Asher Metzger]

Hi there,

In the files attached I am solving a water network with 2 sets of variable.

One of the constraints is bi-linear: Salinity and flow(C and Q).

My thesis is that the larger the problem is the larger the computational time is, in an exponential rate.

These are not the results I am getting. I am actually getting results that seem to be random.

How is this possible when I am solving the network each time with more variables?

I am solving it with Ipopt.

Thanks in advance,

Asher

[Johan Löfberg]

Well, your thesis is wrong then. Nonconvex problems are notoriously unpredictable.

[Asher Metzger]

I love your answer.

I'm just stuck with about a years work now :-)