Probabilistic Dynamic Line Rating Forecasting with Line Graph Convolutional LSTM

TL;DR

The paper addresses the challenge of probabilistic dynamic line rating (DLR) forecasting under weather uncertainty by proposing a spatio-temporal model that combines line-graph convolutional networks with LSTM, augmented by a quantile layer to output prediction intervals. The core method, double-hop line graph convolutional LSTM (D-LGCLSTM), captures extended spatial correlations while mitigating feature duplication and reducing model parameters. Empirical results on the Texas 123-bus backbone show that D-LGCLSTM achieves superior reliability and sharpness (lower ACE, PINAW, IS, QS) with significantly fewer parameters than competitive baselines. The approach promises practical impact for grid operators by enabling tighter, more dependable DLR forecasts and facilitating network-wide, data-driven decision making under uncertainty.

Abstract

Dynamic line rating (DLR) is a promising solution to increase the utilization of transmission lines by adjusting ratings based on real-time weather conditions. Accurate DLR forecast at the scheduling stage is thus necessary for system operators to proactively optimize power flows, manage congestion, and reduce the cost of grid operations. However, the DLR forecast remains challenging due to weather uncertainty. To reliably predict DLRs, we propose a new probabilistic forecasting model based on line graph convolutional LSTM. Like standard LSTM networks, our model accounts for temporal correlations between DLRs across the planning horizon. The line graph-structured network additionally allows us to leverage the spatial correlations of DLR features across the grid to improve the quality of predictions. Simulation results on the synthetic Texas 123-bus system demonstrate that the proposed model significantly outperforms the baseline probabilistic DLR forecasting models regarding reliability and sharpness while using the fewest parameters.

Figures (5)

• Figure 1: Probabilistic and deterministic DLR forecasting.
• Figure 2: Overall Framework of the proposed D-LGCLSTM.
• Figure 3: An example of reliability and sharpness.
• Figure 4: Heat maps of the average QS for the lines in TX-123BT.
• Figure 5: Probabilistic and robust DLR forecasting for line 123.