Dynamic State Estimation of a High-Order Model of Doubly-Fed Induction Generator Using Unscented Kalman Filter

Abstract

The utilization of renewable energy in power generation has been increasing in recent years, with the use of wind power sources being the most promising solution for sustainable power generation. The doubly-fed induction generator (DFIG) is one of the most commonly used generators in wind power generation applications, as it offers some technical of advantages. However, the increasing penetration of wind power generation poses tremendous technical challenges in power system operation as this can potentially affect system stability, requiring better control and monitoring schemes. Dynamic state estimation (DSE) offers the ability to achieve this purpose. With respect to this, the present paper proposes a DSE framework on a high-order model of DFIG consisting of 18 states. The method uses the unscented Kalman filter (UKF) which provides an accurate estimate of DFIG states under a strong system non linearity present in the wind turbine system. Furthermore, this paper demonstrates the robustness of the proposed method under different faults and noisy conditions. Finally, the paper also extends the use of UKF to estimate the unknown inputs of a DFIG system, such as control references in the rotor-side converter (RSC) and grid-side converter(GSC).

Keywords

• Doubly fed induction generator
• dynamic state estimation
• power system monitoring
• unscented Kalman filter
• wind power generation