A robust parametric method for power harmonic estimation based on M-Estimators

Power harmonic estimation is essential for evolving suitable protection and control strategies of power networks. Parametric approaches exhibit super-resolution and have been gradually applied in practice. However, they are susceptible to the presence of noise in the signals. The analysis accuracy will be abruptly reduced and even ineffective under non-Gaussian or impulsive noise environment. In this paper, a novel approach based on M-Estimators for harmonic estimation is proposed to overcome the above shortcomings. The derivation of Gauss-Newton iteration equations of harmonic frequency, amplitude and phase is presented. Further, the ESPRIT algorithm is also employed to acquire the initial values of harmonic frequencies, which avoids the objective function of M-Estimator to be stuck in local minima and improves the convergence rate of optimization. The numerical simulations and experimental results verify the feasibility and effectiveness of the proposed method.