Adaptive spatiotemporal graph learning for multi-horizon probabilistic wind power forecasting

Dong Hua, Geyu Huang, Peiyi Cui, Suisheng Liu, Bojun Deng, Yuanting Feng · 2026

Read Adaptive spatiotemporal graph learning for multi-horizon probabilistic wind power forecasting by Dong Hua, Geyu Huang, Peiyi Cui, Suisheng Liu, Bojun Deng, Yuanting Feng (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-13473523.txt.

Description

Dong Hua and five colleagues developed in 2026 a forecasting system that predicts wind power across many sites and horizons at once. Grid operators need such forecasts to balance a renewables-rich system. Most methods chase accuracy at single horizons or sites, ignoring site links, uncertainty, and erratic jumps over time. Their graph network joins site weather and turbine data in a graph whose links are reweighted as wind patterns change.

A combined loss function targets point accuracy, calibrated probability forecasts, and smooth movement across horizons. In two years of data from twelve diverse wind farm sites, the model cut root mean square error by up to 13 percent at one to four hours, against leading baselines. Gains above 5 percent held at 24 hours. Calibration error fell, with coverage deviations mostly within 3 percent, and rankings stayed steadier over time. The authors offer the approach as a template for multi-site renewable forecasting.

Catalog record

CREATORDong Hua, Geyu Huang, Peiyi Cui, Suisheng Liu, Bojun Deng, Yuanting Feng
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE58 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13473523

Browse the full library · About the Tower of Babel