Abstract
The state estimation ensures the secure and reliable operation of modern-day power systems by accurate monitoring of its states, voltage magnitudes and phase angles. A robust estimator can detect erroneous measurements, also known as bad-data, and eliminate those automatically during the estimation process. Least measurement rejected (LMR) is one of such robust estimators which has been found significantly successful in dealing with bad-data of various categories. In the LMR algorithm, a certain tolerance needs to be imposed on each measurement and the performance of the estimator mostly depends upon it. Different approaches to tolerance selection have been tried in the literature. The novelty of this paper lies in proposing an original approach of improving the accuracy of LMR by appropriate selection of tolerance. The double-layer selection process assigns a unique number to each measurement, based on the meter quality. LMR, with six different versions of tolerance values, is used from the literature, under different bad-data scenarios, to check whether the proposed tuned LMR (TLMR) performs better than those or not. Besides, some widely used robust estimators are simulated to prove the efficacy and robustness of the proposed approach. It has been observed that the estimation accuracy has been improved significantly than the older versions after being subjected to the proposed mechanism. Moreover, TLMR has successfully outperformed the other estimators while dealing with white noise, bad-data and leverage point scenarios.