Wenlei, A good DC OPF formulation should include real power line flow constraints which, of course, will give you "shadow" prices also known as Lagrangian multipliers. I am not sure if Matpower rundcopf() returns branch Lagrangian multipliers but it is trivial to code it in Matlab.
If your problem uses linear cost functions using dual LP formulation is also trivial but I never used dual LP for this problem since it is not needed. In my personal opinion, using branch loading as a congestion measure is not a good idea but I might be wrong. I think that the real question in this problem is how to handle 0/1 variables representing branch on/off. Good luck, Jovan On Mon, Dec 23, 2024 at 5:00 PM Wenlei Bai <[email protected]> wrote: > Dear Matpower Community, > I try to work on a network topology optimization problem where only > transmission switches are the only binary variables. The objective is to > conduct short-term (up to 5 hours ahead) network reconfiguration to reduce > the 'congestion cost' while maintaining the N-1 contingency. > > I plan to develop my own algorithm to tackle the problem, using an > evolutionary algorithm that would allow me to keep the AC power flow > equations in the optimization. 'Congestion cost' can be quantified using > shadow price, but my understanding is that shadow price can only be > calculated by the dual problem of DC OPF, so maybe I can just use the > device loading to represent the congestion cost in the objective function? > > My question is: > 1. Anyone can share insights on what test system is best for this type of > problem? > 2. Any inputs on this problem or how would you start? > > Best, > Wenlei >
