Skip to content

Author

Jiyue Long

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Open access Aug 2026

Big Data-Driven Short-Term Load Forecasting in Smart Grids: A Spatio-Temporal Dynamic Graph Transformer with Uncertainty-Aware Attention

Short-term load forecasting for modern smart grids must jointly model long-range temporal patterns, correlations across many metering points, and the uncertainty required for operational decision-making. We propose ST-DGT-UA, a spatio-temporal dynamic graph Transformer that (i) embeds heterogeneous covariates with convolutional projections and Time2Vec, (ii) learns a time-varying adjacency matrix from node representations, (iii) couples spatial graph attention with temporal ProbSparse attention for efficient long-sequence modeling, and (iv) outputs multiple conditional quantiles trained with multi-quantile Pinball loss. Experiments on GEFCom2014 and the UCI Electricity Load Diagrams datasets show that ST-DGT-UA achieves MAPE/RMSE of 1.92/145.6 on GEFCom2014 and 2.65/42.3 on UCI, and improves probabilistic quality with Pinball Loss 0.024 and CRPS 0.043. These results indicate that learning dynamic, data-drivenspatial dependencies is beneficial for multi-node load forecasting where correlations evolve over time.

Jing Zhou, Jiyue Long, Xiaohong Long · 0 citations