Solver not giving optimal solution

General
1

Hey

I am working on this linear problem with binary variables maximizing CF. Strangely, the MIP solution is always higher whereas the final reported solution is lower even though the solver status is optimal. When I give a constrained lower limit to the CF, the solution is tuned and it now gives another value, still showing proven optimal solution. However if the lower limit more than the MIP solution previously shown, the problem becomes infeasible.

Any help will be greatly appreciated.

2

Any notable difference in the fixed final solve and the MIP solution usually indicates some numerical issue in the model (e.g. poor scaling). See the following thread:
MIP solution and final solve are different