Optimal Reconfiguration of the Eletric Power Distribution Network with High Photovoltaic Generation and Overvoltage

Resumo

Distribution Network Reconfiguration (DNR) is the most common method employed by distribution network operators to achieve optimal system operation. The proliferation of Renewable Energy Sources (RES), particularly photovoltaic (PV) energy generation, in modern distribution networks has introduced novel challenges in network planning and operation. In this context, this paper introduces a DNR strategy involving the manipulation of network topology through switch opening and closing, integrated with a Modified Flower Pollination Algorithm (FPA-M). The primary aim is to minimize active power losses and regulate voltage levels at distribution system buses, particularly under high PV generation. The problem formulation takes the form of a Mixed Integer
Nonlinear Programming (MINLP), encompassing continuous and discrete variables. The FPA-M method is an efficient meta-heuristic algorithm inspired by flower pollination processes. Compared to alternative methods, the FPA-M boasts a single control parameter and proves highly effective in optimization problems. The study’s central contribution lies in enhancing algorithm performance and reducing computational effort through a refined determination of the search space. The proposed approach is evaluated across diverse operational scenarios, including load curves and PV data collected from the Federal University of ABC. Tests are conducted on a 33 bus distribution system, and results are benchmarked against other methodologies from the literature. The findings underscore that the amalgamation of FPA-M with DNR yields substantial energy loss reductions, effectively maintaining voltage levels within acceptable bounds throughout the evaluation period.