Power System State Estimation Under Incomplete PMU Observability-A Reduced-Order Approach

This paper presents a state estimation method for power systems when not all state variables are observable with phasor measurement units (PMU), namely, incomplete PMU observability. Realizing that PMU measurements are generally more accurate than conventional ones, the proposed approach estimates PMU unobservable states and PMU observable states separately. The latter are estimated from PMU measurements using a linear estimator. Those estimates are then used along with conventional measurements in a reduced-order nonlinear estimator for the PMU unobservable states. The proposed decoupled approach features reduced computational complexity and greater numerical stability, when compared to existing combined approaches. We show analytically that the performance of the proposed method is comparable to that of existing approaches if PMU measurements are sufficiently accurate. In this paper, we also study the impact of time-skew errors in conventional measurements on the performance of hybrid state estimators using both conventional and PMU measurements. We propose a model that incorporates such errors into conventional measurements. We show that our proposed method is more robust than existing approaches in the presence of time-skew errors. Our analytical findings are verified by simulations on standard IEEE test systems.