Dynamic Matrix Completion Based State Estimation in Distribution Grids

Abstract

The power distribution network is undergoing tremendous transformation due to an increase in the penetration of renewable energy resources and electric vehicles. These changes have resulted in greater uncertainty and dynamics in the distribution grid states. Therefore, the ability to track and monitor system states has become a critical need for accurate and timely control actions. In this article, we propose two dynamic sparsity-based state estimation approaches for distribution systems: 1) locally weighted matrix completion (LW-MC), and 2) Bayesian matrix completion with Kalman filter prediction (BMC-KF). The performance of the proposed dynamic state estimation strategies is compared with the classic/static matrix completion (static-MC) approach using the IEEE 37 and IEEE 123 bus test systems. Results indicate that BMC-KF approach outperforms both LW-MC as well as static-MC even when 30% of the measurement data is available. Computational complexity associated with both approaches is quantified.