No, there is no mathematical reason. It is simply that it is straightforward to specify general linear constraints as additional data (a matrix and a vector or two) to a solver. On the other hand, facilitating user-specified general non-linear constraints typically requires the ability to provide code for function and derivative evaluations. While this is relatively straightforward for Matlab-based solvers, it was not practical in the case of the MEX-based solvers that were our "best" OPF solvers for many years. So it was simply a matter including some easy-to-implement flexibility into the OPF formulation ... linear was easy, so it was included. Non-linear was way to complicated so it wasn't.
Now that we have a decent pure-Matlab OPF solver (MIPS), it is probably worth exploring the possibility of expanding the options for user specified costs and constraints to something even more general for some future version of MATPOWER. -- Ray Zimmerman Senior Research Associate 211 Warren Hall, Cornell University, Ithaca, NY 14853 phone: (607) 255-9645 On May 19, 2011, at 5:30 AM, Janek Massmann wrote: > Dear all, > > im working on a security constrained optimal power flow. At the moment I am > using add_userfunction to add constraints, which include the outage of > generator and branches. > My problem is the adding of non-linear constraints. Because I can only add > linear constraints I am working with real power and I am neglecting the > apparent power. Because of this effect I cant calculate the absolute workload > of the remaining supplies. Is there any mathematic explanation why I cant add > non-linear constraints? > > Kind regards, > Janek Massmann >
