DYNAMIC BAYESIAN NETWORK FOR WEATHER FORECAST AND EVALUATION OF RENEWABLE RESOURCES AVAILABILITY
In one sentence. Bayesian network learned from four years of station data forecasts wind speed and solar irradiance classes for renewable energy resource evaluation.
Abstract
The authors present a Bayesian network capable to estimate the weather parameters related, not only, to renewable resources: wind speed and solar irradiation. A large and systematic data base about simple and composed weather indices registered during four years, 2013-2016, was used to construct the data-driven Bayesian structure and to learn and validate its parameters. It includes 9 weather indices collected, minute by minute, by a professional Davis Instrument Pro 2 Plus weather station. The extremely large initial data base, over 1.8 million records, was discretized in 4 classes making possible to use a very simple algorithm like Bayesian search to establish the most suitable network structure fitting the data. The main and first useful results mean the probability of wind speed and solar irradiance classes. Both parameters can be transformed in electrical power considering a given wind generator and a solar panel.
Labels
- Topic
- Local & station forecasting, Applications & impacts
- Type
- New model or method
- Method
- Classical ML (trees, GP, linear, kernels)
- Domain
- Local / station
- Data
- Station / in-situ
- Status
- Research
- Identifiers
- OpenAlex:W4289233296
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.