Explored 0 nodes (0 simplex iterations) in 0.00 seconds
ユーザーの入力を待っています。Hi, community!
There is a problem when I use matlab and groubi to solve the optimization problem. I am a new learner, so I am afraid to make some mistakes. So far I think the constraints, objective, and variables are right. But the matlab respones that it processes 0 iteration and the model is infeasible:
and I used gurobi_iis() to detect what constraint is infeasible and obtain the lp document.
I have read the document about infeasible or unbounded model, but I still don't konw how to solve the problem that it doesn't works, so I hope that you can give me some help, and I will be grateful.
These are the constraints:
-
-
Gurobi Staff
Hi,
You can use the MATLAB gurobi_write function to write your model to a human readable LP file. You can then analyze this file by opening it in any common text editor to check whether your model actually looks like what you would expect it to be. Note that you can also write the IIS computed by gurobi to a file by using the \(\texttt{.ilp}\) extension.
You could also comment out parts of your constraints in order to check whether your model becomes feasible. This way you could find which set of constraints is responsible for infeasibility.
Best regards,
Jaromił0 -
-
Thank you for your instruction! But I have another question about my feasible model. Now my model is feasible which is confirmed using MATLAB gurobi_iis().
But now the model can't be solve in an acceptable time. The massage shows that the presolve model has 24 quadratic constraints and root relaxation presolve model has 24 second-order cone constraints.
but in my code, I think my constraints are all linear equation and inequation, so I want to konw why presolve model generates those nonlinear-constraints which I think maybe relate to the long long time that the solver spent to solve my problem.
This is the massage that gurobi returns, which I think helpful to solve my question.
The time is longer than it shows in the picture:
0 -
-
-
From your screenshots, I cannot tell which objects are variables and which are constants. You can have a look at the file generated via the MATLAB gurobi_write function and search for multiplications \(\texttt{*}\) or square terms \(\texttt{^2}\).
At a first glance it looks like part of your constraints are modeling KKT conditions or some convexity conditions which are nonconvex in general.
If it's possible, you could share the LP file you generated as described in Posting to the Community Forum.
0 -
サインインしてコメントを残してください。
コメント
3件のコメント