A photovoltaics-aided interlaced extended Kalman filter for distribution systems state estimation

Abstract Distribution system state estimation (DSSE) is essential for smart grid monitoring and control. Bus voltage phasors and, consequently, DSSE uncertainty can be significantly affected by photovoltaic (PV) penetration, even when suitable hosting capacity strategies are adopted to keep voltage levels within given limits. In this paper, it is shown that the state estimation uncertainty achievable with algorithms exploiting PV information can be significantly lower than using classic techniques, such as the Weighted Least Squares approach that is still widely adopted at the distribution level. The proposed analysis is based on on an Interlaced Extended Kalman Filter (IEKF) that, in the prediction step, relies on the available information on active and reactive power injections. The use of an interlaced implementation makes the estimator more robust to zero-power injections, which otherwise could make the Kalman innovation matrix ill-conditioned. In the update step, the PV power data measured on the field, possibly supported by Phasor Measurement Units (PMUs), complement the virtual measurements, the traditional pseudo-measurements, and those obtained by aggregating smart meter data. The results of one-year-long simulations confirm the benefits of including the available information on PV generation on state estimation uncertainty.