Ok, here are a few more ideas … Did you try checking the bus voltage limits and generator real and reactive power limits for the successful power flow solution? You might find the check_feasibility <http://www.pserc.cornell.edu//matpower/docs/ref/matpower5.1/extras/misc/check_feasibility.html> and checklimits <http://www.pserc.cornell.edu//matpower/docs/ref/matpower5.1/extras/misc/checklimits.html> functions to be useful. If any of these are violated it could point you toward a source of infeasibility. If not, maybe there is some numerical issue affecting the OPF. Try different solvers if you have access to them. You could also try turning on the opf.init_from_mpc option (assuming you are using MIPS, Knitro or Ipopt as your OPF solver) and using the solved power flow as input for the OPF.
Ray > On Jan 7, 2016, at 2:22 PM, Alanazi, Falah <alana...@oregonstate.edu> wrote: > > Dear professor Ray, > > > I eliminated the branch flow limits but still the solution didn't converge. > > also I checked the mpc values and the limits it seems ok. In addition, I > tried to change some of the limits but still did not work > > I attached my data in case that may be helpful to find the problem. > > > On Tue, Jan 5, 2016 at 1:30 PM, Ray Zimmerman <r...@cornell.edu > <mailto:r...@cornell.edu>> wrote: > There are many possible reasons for a non-convergent OPF. One of the first > things I always try is to eliminate the branch flow limits. E.g. > > define_constants; > mpc= loadcase ('WECC179_3Area.mat');. > mpopt = mpoption('out.lim.all',2) > mpc.branch(:, RATE_A) = 0; > results = runopf(mpc, mpopt); > > If that converges, then you likely have an infeasible problem, where some > branch flow limit can’t be satisfied without violating something else. In > that case, you can try multiplying the limits by some large factor that you > gradually reduce toward 1 and observe which constraints (including voltage, > reactive generation) bind hardest. They may show you which are your > conflicting constraints. > > Hope these ideas help point you in the right direction ... > > Ray > > >> On Jan 5, 2016, at 12:42 PM, Alanazi, Falah <alana...@oregonstate.edu >> <mailto:alana...@oregonstate.edu>> wrote: >> >> Dear Ray, >> >> I was running AC PF and it was working but when I tried to run AC OPF the >> solution did not converge. >> This is my code. >> >> define_constants; >> mpc= loadcase ('WECC179_3Area.mat');. >> mpopt = mpoption('out.lim.all',2) >> results = runopf(mpc, mpopt); >> >> best regards >> >> Falah >> >> >> >> >> On Tue, Jan 5, 2016 at 5:40 AM, Ray Zimmerman <r...@cornell.edu >> <mailto:r...@cornell.edu>> wrote: >> The options you are using should work when running an AC OPF. Is that what >> you are running? >> >> Two other notes, regarding your options … >> 1. The ENFORCE_Q_LIMS option is only for power flow, not OPF, and that the >> VERBOSE option only affects display of solution progress, not final results. >> 2. You are mixing old-style option names (all caps) in your first line, with >> new-style options in the second. This still works, but the old-style options >> are deprecated. The new versions of ENFORCE_Q_LIMS and VERBOSE are >> pf.enforce_q_lims and verbose, respectively. >> >> — Ray >> >> >> >>> On Jan 4, 2016, at 5:09 PM, Alanazi, Falah <alana...@oregonstate.edu >>> <mailto:alana...@oregonstate.edu>> wrote: >>> >>> Dear Sir, >>> >>> I would like to know how I can show the voltage constraints and the >>> branch flow constraints when I run OPF. >>> >>> When I used the matpower cases both constraints showes up but when I used >>> the WECC data they do not show up. I tried to use >>> >>> mpopt = mpoption('ENFORCE_Q_LIMS',0,'VERBOSE',3); >>> mpopt = mpoption(mpopt,'out.lim.all',2) >>> >>> best regards >>> >>> Falah >> >> > > > <WECCPFlow.m>
