Conference article

Wind Speed Prediction based on Incremental Extreme Learning Machine

Elizabeta Lazarevska
Faculty of Electrical Engineering and Information Technologies – Skopje, University “Ss. Cyril and Methodius” – Skopje, Macedonia

Download articlehttp://dx.doi.org/10.3384/ecp17142544

Published in: Proceedings of The 9th EUROSIM Congress on Modelling and Simulation, EUROSIM 2016, The 57th SIMS Conference on Simulation and Modelling SIMS 2016

Linköping Electronic Conference Proceedings 142:79, p. 544-550

Show more +

Published: 2018-12-19

ISBN: 978-91-7685-399-3

ISSN: 1650-3686 (print), 1650-3740 (online)

Abstract

There are many research papers dealing with wind speed forecasting, since it is necessary in many applications, such as agriculture, modern transportation, and wind energy production. This paper presents an alternative approach to modeling and prediction of wind speed based on extreme learning machine, which is gaining a considerable interest in the scientific and research community at the present. Since the wind speed depends on the atmospheric weather conditions, the wind speed forecast in this research is based on different meteorological data, such as ambient temperature, relative humidity, light intensity, dew point, and atmospheric pressure. The paper presents two neural models for wind speed prediction, based on classic and incremental extreme learning machine, which exhibit the attributes of extreme simplicity, extremely good approximation performance, and extremely fast computation. The performance of the models is validated through their performance indices and compared to other available fuzzy and neural models for wind speed prediction. The paper also addresses the applied modeling techniques and proposes a modification which gives improved results and better approximation performance than the original techniques.

Keywords

wind speed prediction, extreme learning machine, incremental extreme learning machine, random nodes

References

No references available

Citations in Crossref