Simultaneous Dynamic State and Parameter Estimation of DFIG With Unknown Inputs

This paper presents an Unscented Kalman Filter (UKF) algorithm for state estimation in a Doubly Fed Induction Generator (DFIG) wind turbine model, emphasizing the generator’s crucial role as a source of electrical energy in maintaining power system stability. Given the significance of monitoring the generator state, dynamic state estimation becomes imperative to assess system condition and ensure power system stability. Furthermore, parameter estimation is essential in DFIG due to the limitations of constant parameters in accurately representing the system’s dynamic behavior. To address these challenges, this study introduces a dynamic state estimation (DSE) framework for the DFIG model, comprising 18 states. Additionally, estimation procedures for both inputs and parameters are incorporated to enhance representation fidelity of the generator’s actual condition. Results demonstrate the UKF method’s reliability in estimating state, input, and parameters under various disturbance and noise conditions.