ROBUST STATE ESTIMATION IN POWER SYSTEMS

Abstract

Precise state estimation is a prerequisite to effective management of power systems. However, there are still issues in the robustness of power system state estimation. Some of these issues such as the optimization of robust state estimators in power systems are studied in the thesis. Optimal placement of phasor measurement units is also considered. Not only the minimum number of phasor measurement unit and maximum measurement redundancy but also the minimum variance of the state estimation error can be obtained. Traditionally power systems are assumed to be at steady-state and for simplicity, state estimation was designed for a steady-state system. However, under normal circumstances, the states change slowly and continuously. To capture their dynamic behaviors and monitor the system, a new recursive t-distribution based maximum likelihood estimator for dynamic power systems is proposed, which is robust to outliers.