Restauração de redes de distribuição inteligentes sob incerteza

Abstract

Currently, there is a high level of connection of Photovoltaic Distributed Gen- eration (PVDG) in distribution networks. The presence of PVDG affects the restoration of the power distribution network, since PVDG can provide active and reactive power support to mitigate violations of voltage and loading con- straints during load transfers via normally open switches to adjacent feed- ers.The post-reconfiguration state, obtained by the restoration plan, is subject to the risk of violations of technical constraints associated with voltage and/or loading limits. This risk is caused by uncertainties in the PVDG output power. Another source of uncertainty that can increase the risks of violations of tech- nical constraints during restoration are the fluctuations in the active and re- active loads at each node due to errors in the measurement system. This dis- sertation presents a methodology to perform the restoration of radial distribu- tion networks subject to uncertainties associated with the load and PVDG. These uncertainties were modeled in a probabilistic power flow algorithm, based on the Latin Hypercube, which calculates the risks of voltage and load constraint violations for a candidate topology for restoration. The restoration problem was solved using the Tabu Search algorithm. This method was com- bined with a Fuzzy structure to consider a multi-objective restoration based on the following criteria: restored active power and risk of technical constraint violation. The test results demonstrated that the deterministic restoration re- sults in high risks of constraint violations in the optimal solution. On the other hand, the proposed multi-objective technique obtained an acceptable compro- mise between the risk of violation and the restored active power.