Solar radiation prediction using multivariate signal decomposition and physics-informed time-frequency feature extraction
Xingchen Mo, Jingzhou Xin, Yan Jiang, Qizhi Tang · 2026
Read Solar radiation prediction using multivariate signal decomposition and physics-informed time-frequency feature extraction by Xingchen Mo, Jingzhou Xin, Yan Jiang, Qizhi Tang (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-13351062.txt.
Description
Xingchen Mo and co-authors forecast sunlight the way a signal engineer would take apart a chord. Solar power planning needs accurate radiation forecasts, yet the signal jumps with weather in tangled patterns. Their method first splits records from many stations at once into aligned layers of rhythm, keeping the links between stations and between rhythms.
A custom neural network then reads those layers, using attention guided by two physics measures of how rhythms lock together. The point is to let physical coherence steer the learning instead of leaving the network to guess from raw curves. In tests on measured multi-station data, the method beat rival approaches across several forecasting settings.
Catalog record
| CREATOR | Xingchen Mo, Jingzhou Xin, Yan Jiang, Qizhi Tang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |