Probability density function selection based on the characteristics of wind speed data

Yürüsen, N.Y. ; Melero, Julio J. (Universidad de Zaragoza)

Resumen: The probabilistic approach has an important place in the wind energy research field as it provides cheap and fast initial information for experts with the help of simulations and estimations. Wind energy experts have been using the Weibull distribution for wind speed data for many years. Nevertheless, there exist cases, where the Weibull distribution is inappropriate with data presenting bimodal or multimodal behaviour which are unfit in high, null and low winds that can cause serious energy estimation errors. This paper presents a procedure for dealing with wind speed data taking into account non-Weibull distributions or data treatment when needed. The procedure detects deviations from the unimodal (Weibull) distribution and proposes other possible distributions to be used. The deviations of the used distributions regarding real data are addressed with the Root Mean Square Error (RMSE) and the annual energy production (AEP).

Idioma: Inglés

DOI: 10.1088/1742-6596/753/3/032067

Año: 2016

Publicado en: Journal of physics. Conference series 753 (2016), 032067 [10 pp.]
ISSN: 1742-6588

Factor impacto SCIMAGO: 0.24 - Physics and Astronomy (miscellaneous) (Q3)

Financiación: info:eu-repo/grantAgreement/EC/H2020/642108/EU/Advanced Wind Energy Systems Operation and Maintenance Expertise/AWESOME

Tipo y forma: Artículo (Versión definitiva)

Área (Departamento): Área Ingeniería Eléctrica (Dpto. Ingeniería Eléctrica)