Semicentralized voltage control method using SVR based on past voltage measurements in distribution network

In this paper, we propose a semicentralized voltage control method utilizing past voltage measurements obtained with sensors in a distribution network (DN) to improve the voltage control performance of a single-step voltage regulator (SVR). In this method, the SVR basically performs autonomous voltage control referring to the control parameter, the reference voltage, but the parameter is updated with a specific interval via communication. To determine a time series of the reference voltage using the past voltage measurements, the simple averaging method and the probabilistic weighting method are considered. The effect of these methods on the voltage control performance is evaluated with a power flow simulation using a realistic 6.6 kV DN model with photovoltaic systems. The results show that the proposed semicentralized SVR control with these advanced parameter-updating methods dramatically mitigate voltage violation compared to the conventional SVR operation in which the constant reference voltage is used for 1 year.