Interior point methods tend to scale quite well for large, sparse optimization problems such as the OPF.
-- Ray Zimmerman Senior Research Associate B30 Warren Hall, Cornell University, Ithaca, NY 14853 phone: (607) 255-9645 On Jun 19, 2014, at 12:28 PM, Abhay Kumar <[email protected]> wrote: > I am curious about the reason Matpower uses interior point methods for AC > OPF. I am new to optimization and very impressed by the numbers that Dr. > Zimmerman just shared (about the system size). > > An expert opinion would be very helpful. > > Regards > > Abhay > > > On Thu, Jun 19, 2014 at 11:54 AM, Ray Zimmerman <[email protected]> wrote: > There is no explicit limit, so the size will be limited by RAM and/or > computation time. I have successfully solved an AC OPF problem with ~193,000 > buses, ~25,000 generators, ~230,000 branches with the default interior-point > solver. > > -- > Ray Zimmerman > Senior Research Associate > B30 Warren Hall, Cornell University, Ithaca, NY 14853 > phone: (607) 255-9645 > > > > On Jun 19, 2014, at 10:22 AM, Abouzar Estebsari <[email protected]> wrote: > >> Dear all >> >> I would like to find out the allowed dimension of MATPOWER input cases >> including the number of bus, branch and generators. As an example, I want to >> know how large the input network can be, especially how many generators >> including dispatch-able loads can be handled by the default interior-point >> method solver provided by MATPOWER. >> >> Did anybody tried? >> >> Thanks a lot, >> >> Abouzar >> >> -- >> Abouzar Estebsari >> Doctoral Researcher >> <image001.jpg> >> Dipartimento Energia >> Corso Duca degli Abruzzi, 24 >> 10129 Torino – Italy >> tel. +39 011 090 7117 >> fax +39 011 090 7199 >> > >
