Hi Willem,
Have you checked that the power flow solution satisfies all of the OPF
constraints? I’m thinking specifically the voltage, branch flow and generator
active and reactive limits. The power flow does not take into account these OPF
limits, and I suspect there are violations. You may find the checklimits()
function in the extras/misc directory to be helpful here. You can use it as
follows:
r = runpf(<your_case>);
checklimits(r);
Hope this helps,
Ray
On Jun 11, 2021, at 11:11 AM, Willem Esterhuizen
<wil.es...@protonmail.com<mailto:wil.es...@protonmail.com>> wrote:
Dear MATPOWER community,
I have an initial power flow solution for an mpc case file obtained via runpf().
If I then invoke runopf() with quadratic or linear cost on power generation the
solver exits without a solution. This is very strange because there exists a
feasible solution which I compute with runpf(), and with which I start runopf().
I have considerably relaxed the bounds on PMAX, PMIN, QMAX, QMIN, VMAX, VMIN on
all entries of mpc.bus and mpc.gen. I have tried FMINCON and MIPS.
Some more information: this phenomenon occurs for the grids from the Simbench
project: https://simbench.de/en/ (at least, I tried it for 5 different grids
and they all gave the same result). The original data is in csv, but available
in pandapower via the simbench package. The OPFs also do not converge in
pandapower. I exported the pandapower instances to Matpower mpc cases to test
in Matlab.
Any help would be greatly appreciated. Please see details of the error messages
below.
Regards,
Willem
My errors in Matlab:
The grid has 1 generator and 102 loads. The initial feasible point for the
generator is:
Bus Voltage Generation Load
# Mag(pu) Ang(deg) P (MW) Q (MVAr) P (MW) Q (MVAr)
----- ------- -------- -------- -------- -------- --------
1 1.025 0.000* -31.63 12.20 - -
If I use the MIPS solver I get:
Warning: Matrix is close to singular or badly scaled. Results may be
inaccurate. RCOND = 1.960342e-29.
Numerically Failed
Did not converge in 16 iterations.
>>>>> Did NOT converge (0.56 seconds) <<<<<
If I use FMINCON I get:
MATPOWER Version 7.1, 08-Oct-2020 -- AC Optimal Power Flow
AC OPF formulation: polar voltages, power balance eqns
Converged to an infeasible point.
fmincon stopped because the size of the current step is less than
the value of the step size tolerance but constraints are not
satisfied to within the value of the constraint tolerance.
<stopping criteria details>
>>>>> Did NOT converge (2.25 seconds) <<<<<
mpc.gen for this infeasible point returned by FMINCON reads:
1.0000 -28.7893 21.6434 300.0000 -300.0000 0.5739 1.0000 1.0000
200.0000 -200.0000
where the entries are: bus Pg Qg Qmax Qmin Vg mBase status Pmax Pmin,
in that order.
