Optimal Reconfiguration of Electric Power Distribution Networks

The study proposes a methodology to address the reconfiguration of electrical distribution networks, formulating it as a mixed integer nonlinear optimization model. The complexity arises from disconnectors, with binary behavior, and nonlinear power flow equations. The lack of efficient and reliable programs to solve these challenges is highlighted. A solvable approach is proposed as a mixed integer convex model, implemented in Julia with JuMP for optimization and the solvers Ipopt and CPLEX. It is compared with the standard operation of the system to validate the proposed formulations, including relaxed models of convex nature. The implementation is validated on two case studies, one created for the research and the other using a 118-bar IEEE radial system. The results show significant advantages in computational speed and response quality, improving the operating cost of the distribution system between 1% and 12%. It is important to highlight that the modification of disconnectors does not imply additional costs in the operation of the system. This work contributes with efficient convex models to solve reconfiguration problems in electrical networks.