PMU-Based Estimation of Synchronous Machines’ Unknown Inputs Using a Nonlinear Extended Recursive Three-Step Smoother

Knowledge of the synchronous machines' control input signals and internal states can provide valuable insight to system operators for assessing security margins and the stability of the power system. For example, during disturbances in a stressed power system, it can be of great value to monitor the performance of the machine's control system, e.g., the response of the field voltage, mechanical power, and the field current. As there are often no real-time power plant measurements available to power system operators, internal states, and unknown inputs of generator units would need to be estimated from synchrophasor measurements. This paper proposes a new estimation algorithm, the nonlinear extended recursive three-step smoother (NERTSS), to simultaneously estimate the states and the unknown inputs of the synchronous machine using data from phasor measurement units. These quantities can then be used to monitor the performance of the machine's controls. The case studies presented in the paper compare the estimation performance of the NERTSS with the extended Kalman filter with unknown inputs (EKF-UI) when noisy synchrophasor measurements are used. The simulation results show that the proposed estimation method compares favorably with respect to the EKF-UI in terms of the achieved estimation accuracy.