Probabilistic Algorithm based on 2m+1 Point Estimate Method Edgeworth considering Voltage Confidence Intervals for Optimal PV Generation

Abstract

Photovoltaic (PV) systems widespread into distribution networks due to its environmentally friendly source of energy, cost-competitive option and system support benefits. However, traditional distribution networks were not designed to operate under a high penetration of intermittent generation posing technical challenges for grid operation and planning. Therefore, probabilistic tools become suitable to cater for uncertainties in generation and demand, thus, leading to a more realistic network representation. Furthermore, the need for harvesting potential energy in an uncertain environment are essential for an efficient grid operation. In this context, this work proposes a probabilistic algorithm based on 2m+1 Point Estimate Method Edgeworth to tackle technical issues considering voltage confidence levels that is used for maximizing PV generation. Tests in a IEEE 33 buses radial distribution system using the proposed probabilistic algorithm yields higher accuracy of cost probability distribution, voltage confidence intervals and a faster computational time when compared to Monte Carlo simulation.