You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I tested PowerModels.jl (the SOCWRPowerModel) on the pg-lib test case pglib_opf_case3_lmbd.m and noticed that the reference bus is not constrained to 1 p.u. during the optimisation but instead to the provided min and max values (e.g. 0.9 and 1.1 respectively).
Therefore, the results differ from the results obtained by PandaModels.jl, since they tighten the min and max values for the reference bus.
Is this behaviour desired?
I am looking forward to a reply and am happy to provide further details if desired.
The text was updated successfully, but these errors were encountered:
Hi @SimonUlm. The default OPF problem specification in PowerModels is consistent with the PGLib-OPF version. This is the standard formulation for algorithmic benchmarking, which is why we prefer it as the default in this package. So, yes, this is the desired behavior.
There are plenty of good reasons one might depart from this default and I am sure PandaPower had one such reason in mind. PowerModels provides an easy framework for developing new problem formulations so that one can explore variants like these with a minimum of additional effort.
I tested PowerModels.jl (the
SOCWRPowerModel
) on the pg-lib test case pglib_opf_case3_lmbd.m and noticed that the reference bus is not constrained to 1 p.u. during the optimisation but instead to the provided min and max values (e.g. 0.9 and 1.1 respectively).Therefore, the results differ from the results obtained by PandaModels.jl, since they tighten the min and max values for the reference bus.
Is this behaviour desired?
I am looking forward to a reply and am happy to provide further details if desired.
The text was updated successfully, but these errors were encountered: