Monitoring of subsynchronous oscillation in a series-compensated wind power system using an adaptive extended Kalman filter

The characteristics of subsynchronous oscillation (SSO) in series-compensated wind power systems are significantly affected by the system's operating condition. Besides the variation in the magnitude and frequency of the SSO during an SSO event, the fundamental frequency may also deviate from its nominal value. This study aims at capturing the dynamics of both subsynchronous and fundamental frequency components simultaneously. This work first explores the strengths and weaknesses of various Kalman filtering based frequency tracking algorithms for joint tracking of the fundamental and SSO components. Then, it proposes a novel adaptive extended Kalman filtering (AEKF) algorithm, in which the process noise covariance is updated online by maximising the probability density function of the predicted error. The process noise covariance factor changes to positive non-zero value whenever the frequency of the fundamental component deviates. Thus, the proposed AEKF extracts the time-varying subsynchronous component while also tracking the small variations in the fundamental frequency. The tracking performance of the AEKF is validated on computer-generated test signals as well as on the electromagnetic transient simulation model of an actual wind power system facing SSO. The captured fundamental and subsynchronous dynamics can be used for designing monitoring, protection, and control schemes for the SSO.