Dynamic Harmonic Estimation Using a Novel Robust Filtering Strategy: Iterated Sliding Innovation Cubature Filter

Active filters (AFs) are effective tools for mitigating the detrimental effects of harmonic components on the power systems. The performance of AFs is significantly dependent on designing an accurate and robust estimator which is responsible for providing reference harmonic values. In this article, a novel technique, called sliding innovation (SI) cubature filter, is proposed to estimate the harmonic parameters, i.e., magnitude and phase, in various operating conditions. The proposed method exploits the concept of sliding mode control in the formulation of the measurement update step in the Bayesian filtering framework to enhance the robustness of the estimator. Furthermore, the iterated version of the proposed algorithm, called iterative SI cubature filter, is presented to enhance the accuracy of the estimator. The proposed method keeps its robustness and accuracy in the noisy conditions under the fault occurrence as well as power system transients. The obtained results from both simulation and experimental setup confirm that the proposed estimator is more accurate and robust with a higher convergence speed compared to the well-known discrete Fourier transform (DFT), cubature Kalman filter (CKF), iterated extended Kalman filter (IEKF), and particle filter (PF).