G-2025-52
Blackbox optimization for loss minimization in power distribution networks using feeder reconfiguration
, , and
BibTeX referenceModern power distribution networks (DNs) increasingly incorporate active distribution network technologies, such as distributed energy resources (DERs) and remotely activated switches. As DNs are naturally unbalanced due to a multi-phase, fluctuating demand, DERs which can lead to bidirectional power flows amplify the phase imbalances, reducing the system reliability and efficiency. The proposed network topology reconfiguration method uses tie and sectionalizing switches to minimizes power losses in a three-phase, unbalanced DN equipped with DERs. Strict feasibility of the solution is ensured through a high-accuracy load-flow simulator and a blackbox optimization (BBO) formulation. To circumvent the computational burden of BBO, combinatorial optimization-inspired algorithms are adapted to the DN context, namely the variable neighbourhood search metaheuristic and the branch-and-bound framework. The methods are tested on the IEEE 34-bus, 136-bus, and IEEE 8500-bus systems, all integrating DERs. Results demonstrate the direct impact of combining local generation with network reconfiguration to improve DN efficiency. Notably, the solution typically results in a topology different from the original one. Moreover, power losses are considerably reduced across all test cases, with decrease of at least 36.94% for the largest test system and 9.82% for the practical IEEE 8500-bus case. The results also permit to identify the most suitable methods for practical deployments based on prioritized requirements.
Published July 2025 , 12 pages
Research Axis
Research applications
Document
G2552.pdf (1000 KB)