HAL : hal-00137536, version 1

DOI : 10.1016/S0960-1481(03)00019-3

Fiche détaillée

Récupérer au format BibTeX - EndNote - TEI - RefWorks -

Renewable Energy 28, (2003) 1787-1802.

Statistical bivariate modelling of wind using first-order Markov chain and Weibull distribution.

F. Youcef-Ettoumi 1, H. Sauvageot 2, A.-E.-H. Adane 1

(2003)

This paper studies the statistical features of the wind at Oran (Algeria). The data used are the wind speed and wind direction measurements collected every 3 h at the meteorological station of Es Senia (Oran), during the 1982/92 period. The eight directions of the compass card have been considered to build the frequency distribution of the wind speed for each month of the year and each direction. The three-hourly wind data have been modelled by means of Markov chains. First-order nine-state Markov chains are found to fit well the wind direction data, whereas the related wind speed data are well fitted by first-order three-state Markov chains. The Weibull probability distribution function has also been considered and found to fit the monthly frequency distributions of wind speed measurements. Two methods of wind data retrieval are thus made available. In fact, two models of chronological bi-series are obtained describing wind speed and wind direction.

1 :	Université des Sciences et de la Technologie HOUARI BOUMEDIENE (USTHB)
	Université des Sciences et de la Technologie Houari Boumediène
2 :	Laboratoire d'aérologie (LA)
	CNRS : UMR5560 – Observatoire Midi-Pyrénées – INSU – Université Paul Sabatier [UPS] - Toulouse III

Domaine

:

Physique/Physique/Physique Atmosphérique et Océanique

Mots Clés : Wind modelling – Stochastic Processes – Markov chain – Weibull distribution – Long-term probability

hal-00137536, version 1

http://hal.archives-ouvertes.fr/hal-00137536

oai:hal.archives-ouvertes.fr:hal-00137536

Contributeur : Maria-Isabel Ledig <>

Soumis le : Mardi 20 Mars 2007, 14:58:49

Dernière modification le : Lundi 12 Septembre 2011, 13:00:31