Distributed self supervised method improves energy forecasting accuracy
Distributed JEPA: A Self-Supervised Framework for Energy Forecasting
Machine LearningArtificial Intelligence
Summary
Forecasting energy use or production usually needs lots of specific data and tailored models. The authors created a new approach called distributed JEPA that learns general patterns from many different kinds of energy data without needing labeled examples. This approach predicts hidden parts of the data using shared representations that keep information varied and stable. Their tests show that JEPA forecasts energy consumption and solar power generation accurately and handles missing data better than a common Transformer method.
What this means in practice
- •For energy system operators: Improve robustness and transferability of energy consumption forecasts across heterogeneous buildings without task-specific training data.
- •For solar farm managers: Obtain more accurate and stable forecasts for unseen solar photovoltaic installations even with incomplete data inputs.
Authors
Liana Toderean, Tudor Cioara, Vasilis Michalakopoulos, Efstathios Sarantinopoulos, Ionut Anghel, Elissaios Sarmas
Abstract
Traditional energy forecasting solutions rely on task-specific supervision and energy asset representations, limiting transferability and the ability to capture general temporal dynamics across heterogeneous assets. We address this by proposing a distributed Joint Embedding Predictive Architecture (JEPA) for self-supervised learning from heterogeneous energy time-series. The framework predicts latent representations of masked temporal segments while integrating temporal observations and contextual information within a shared embedding space. To prevent representation collapse, training combines a latent-space predictive objective with covariance and temporal variance regularization. The evaluation was conducted on energy consumption and generation datasets under data-degradation scenarios and compared with a Transformer forecasting baseline. The learned representations remained stable (cosine similarity $\approx 0.98$; effective rank 185-235). JEPA achieved performance comparable to a Transformer on building energy data, higher $R^2$ in 3/5 consumer clusters, and outperformed the baseline on 9/10 unseen PVs ($R^2$=0.73-0.88 vs. <0.45), while showing greater robustness to missing data.