An ML-based ERA Algorithm for Estimation of Modes Utilizing PMU Measurements

This paper proposes a Machine learning (ML) based Eigen Realization algorithm (ERA) for real-time estimation of poorly damped modes in power system. The proposed technique is designed in such a way that it mitigates the effects of data loss and presence of outliers from the Phasor Measurement Unit (PMU) which occurs because of communication delay, hardware faults etc. There is an enhancement in ERA algorithm where a fast-computing weighted moving average (WMA) and inter-Quartile range (IQR), combinedly called WMA-IQR filter has been introduced to take care of the incomplete measurements and outliers respectively present in the system. Finally, the ERA algorithm is applied using a robust data set to give a precise value of the modes. The robust performance of the proposed scheme over the ERA and modified ERA estimator has been illustrated on some simulations carried out on synthetic signals and real signal collected from WECC.