opf-admm

Distributed Optimal Power Flow with ADMM via Message Passing Interface (MPI) implemented in C++.

Report (Spanish)

Based on Q. Peng and S. H. Low, "Distributed Optimal Power Flow Algorithm for Radial Networks, I: Balanced Single Phase Case," in IEEE Transactions on Smart Grid, vol. 9, no. 1, pp. 111-121, Jan. 2018.

Requirements

• Matlab
• CVX
• mpich

Setup

For the setup you need to:

• edit the path to the opf-admm directory in the file NetworkModel.hpp
• edit the path to the matlab binaries, i.e "/usr/local/MATLAB/R2017a/bin/matlab" in files m/update_x.sh and m/update_y.sh
• for each node you can modify the SOCP optimization problem, editing the files 0,1...,7/update_x.m and 0,1...,7/update_y.m

The default program works with 8 buses and 8 MPI processors, this can be configured changing the file NetworkModel.hpp and adjusting the makefile.

Compiling

make

Run

make run

Clean

You can delete previous data with:

make clean

Example

8-bus configuration example:

Active power injected: