Refined convolution-based measures for real-time harmonic distortions estimation in power system dominated by inverter-based resources

Real-time estimating harmonic distortion levels near the Point of Interconnections of grid-connected inverter-based resources (IBRs) is critical for maintaining the operational reliability and security of the power grid, particularly during transients when IBRs generate excessive harmonics. Traditional methods like Total Harmonic Distortion (THD) are unsuitable for time-varying waveforms as they suffer from aliasing, spectral leakage, and picket-fence effects. This paper proposes a novel approach that utilizes convolution-based measures for estimating harmonic distortion, which offers a more suitable way for ongoing real-time applications. The first measure uses a 60 Hz sinusoidal function as the convolution kernel, while an enhanced measure employs the Gaussian function as the kernel. Two convolution-based measures are proposed and compared with THD. The comparison criteria are based on the two significant features describing the variation of harmonics: min/max values and variation curvature of harmonics showing the accurate performance of proposed measures. Moreover, short-term estimations of harmonic distortion are used to validate the effectiveness of the proposed measures, while long-term estimations are used to determine the ability of the proposed measures to determine harmonic distortion contributions among power network components. The proposed technique's intrinsic characteristics provide a sampling window with low sensitivity to deviations in the fundamental frequency and signal's stationary condition to prevent aliasing and spectral leakage and rule out the impact of the picket-fence effect on the harmonic distortion level estimation. This paper presents a novel convolution-based approach for real-time estimation of harmonic distortion levels in power networks, addressing the challenges faced by traditional methods such as Total Harmonic Distortion (THD). The proposed method demonstrates superior performance in terms of accuracy, computational efficiency, and sensitivity to deviations in fundamental frequency, offering an effective alternative for monitoring harmonic distortions in steady-state and transient-state scenarios.image