A new method of wind speed prediction based on weighted optimal fuzzy c-means and modular extreme learning machine
| Autor: | Qide Tan, Benshuang Qin, Chao Pan |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Renewable Energy
Sustainability and the Environment business.industry Computer science 020209 energy Energy Engineering and Power Technology 02 engineering and technology Modular design Fuzzy logic Wind speed Control theory Physics::Space Physics 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Volatility (finance) business Physics::Atmospheric and Oceanic Physics Randomness Extreme learning machine |
| Zdroj: | Wind Engineering. 42:447-457 |
| ISSN: | 2048-402X 0309-524X |
| DOI: | 10.1177/0309524x18779337 |
| Popis: | According to the characteristics of randomness, volatility, and unpredictability of wind speed, this article provides a new wind speed prediction method which includes three modules that are attribute weighting module, intelligent optimization clustering module, and wind speed prediction module based on extreme learning machine. First, the Pearson coefficient values of the attribute matrix elements are calculated and weighted considering the fluctuation characteristics of time series and influences of different weather attributes on the wind speed. Then the fuzzy c-means clustering method optimized by genetic simulated annealing algorithm is carried out on the weighted attribute matrix to cluster. Furthermore, several kinds of wind speed prediction models are built using the extreme learning machine to forecast short-term wind speed. The research on wind speed prediction is carried out by the measured data of wind farm in America (N39.91°, W105.29°). And the results show that the new prediction method of wind speed proposed in this article has higher prediction accuracy. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|