Explainable attention-based neural network for load forecasting in super smart grids using socioeconomic and power consumption data
José Gerardo Silos García, Israel Macias, Ricardo Enrique Gutiérrez Carvajal · 2026
Read Explainable attention-based neural network for load forecasting in super smart grids using socioeconomic and power consumption data by José Gerardo Silos García, Israel Macias, Ricardo Enrique Gutiérrez Carvajal (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13385810.txt.
Description
José Gerardo Silos García and co-authors forecast electricity demand for future super smart grids with an explainable neural network, in a 2026 paper. Planning such grids needs demand estimates that use social data as well as past consumption, since older methods lean too hard on one region's time series. Their model runs two networks in parallel over the two data streams and fuses them with an attention mechanism.
Tested on ninety-two Australian suburban zones, the method cut forecast error to under two percent, beating standard recurrent models when social metrics were included in training. Two explanation tools then reveal which inputs drove each forecast. The authors link better, fairer forecasts to pricing and supply choices that serve all zones.
Catalog record
| CREATOR | José Gerardo Silos García, Israel Macias, Ricardo Enrique Gutiérrez Carvajal |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 135 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |