Probabilistic Ampacity Forecasting of Dynamic Line Rating Considering TSOs Risk-Averse

Abstract

High penetration of renewable energy resources with highly probabilistic loading in the emerging power transmission network is forcing Transmission System Operators (TSOs) to utilize their resources to the exhaustive extent by making use of intelligent transmission network management methods. The real-time ampacity of overhead conductors is tremendously fluctuating due to its dependence on weather conditions. As a result, the real-time rating of the overhead conductor is better exploited by using dynamic line rating (DLR) than traditional conservative static rating, which depends on the worst-case weather conditions. Since there are high uncertainties associated with point forecast DLR ampacity calculation, probabilistic means of DLR forecasting method provide the possibility for short-term planning and real-time overhead transmission line ampacity monitoring, thus enabling the transmission network to run smoothly without harm to the entire network. In this study, a real-time DLR overhead transmission line is formulated, giving 24-hour ahead ampacity prediction and loading limits by using quantile regression forest (QRF) machine learning model with different quantiles. The proposed method provides better enhancement and safe operation for the lowest quantiles to mitigate decision-makers risk-averse.