A Mixed-Integer Nonlinear Programming Model for Optimal Reconfiguration of DC Distribution Feeders

O. D. Montoya,W. Gil-González,J. C. Hernández,D. A. Giral-Ramírez,A. Medina-Quesada
DOI: https://doi.org/10.3390/en13174440
IF: 3.2
2020-08-27
Energies
Abstract:This paper deals with the optimal reconfiguration problem of DC distribution networks by proposing a new mixed-integer nonlinear programming (MINLP) formulation. This MINLP model focuses on minimising the power losses in the distribution lines by reformulating the classical power balance equations through a branch-to-node incidence matrix. The general algebraic modelling system (GAMS) is chosen as a solution tool, showing in tutorial form the implementation of the proposed MINLP model in a 6-nodes test feeder with 10 candidate lines. The validation of the MINLP formulation is performed in two classical 10-nodes DC test feeders. These are typically used for power flow and optimal power flow analyses. Numerical results demonstrate that power losses are reduced by about 16% when the optimal reconfiguration plan is found. The numerical validations are made in the GAMS software licensed by Universidad Tecnológica de Bolívar.
energy & fuels
What problem does this paper attempt to address?