Determination of probabilistic risk of voltage collapse using radial basis function (RBF) network

Abstract

This paper describes a viewpoint for voltage stability assessment accounting uncertainties in line parameters and settings of reactive power control variables. A probabilistic risk of voltage collapse, however small it may be, is always present if system parameters and control variables are treated as random variables. Such uncertainties become important if operating point of system is near to voltage collapse point. Monte-Carlo simulation has been used to evaluate probabilities of voltage collapse for various operating conditions. Static voltage stability limit for various sampled values of system parameters and control variables have been obtained using continuation power flow methodology. Monte-Carlo simulation is a time-consuming process. Hence, a radial basis function (RBF) network has been used to get probabilistic risk of voltage collapse. Training and testing instances have been generated using Monte-Carlo simulation. The algorithm developed has been implemented on two standard test systems.