This code implements a distributed consensus algorithm (DCA) for optimial power flow (OPF) problems using a special Radial Partitioning. The code uses the python modelling package for optimization Pyomo, using the solver Ipopt.
Some case instances from the MATPOWER library that are trasformed into Pyomo data format are given in the Data folder. There is MATLAB m file opf_data_file.m in the Data folder that transforms MATPOWER case files into the Pyomo DAT file for DiCARP.
For using DiCARP, Python, Pyomo, and the required modules must be installed. This code is using Ipopt as the underlying solver, which must be installed as well. The easiest way to do is downloading ipopt.exe from this site and putting it in the required forlder.
The main file has a code to run DiCARP for a given problem instance. The parameters to set are:
problem: A string with the name of the problem, for example"case9".rho: The parameter of the distributed algorithm.tol: The tolerance for the stopping criteria.mat_iter: Maximum number of iterations set for the distributed algorithm.
The main function is DCA_algorithm which returns a python dictionary with the results and statistics of the distributed algorithm.
DiCARP is based on the following manuscript:
Mehdi Karimi. Distributed Optimization for Power Systems with Radial Partitioning, https://arxiv.org/abs/2305.01032, 2023.