An alternative distribution to Weibull for modeling the wind speed data: Inverse Weibull distribution


Akgul F. G., ŞENOĞLU B., Arslan T.

ENERGY CONVERSION AND MANAGEMENT, cilt.114, ss.234-240, 2016 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 114
  • Basım Tarihi: 2016
  • Doi Numarası: 10.1016/j.enconman.2016.02.026
  • Dergi Adı: ENERGY CONVERSION AND MANAGEMENT
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Sayfa Sayıları: ss.234-240
  • Anahtar Kelimeler: Wind speed, Weibull, Inverse Weibull, Monte Carlo simulation, Efficiency, NUMERICAL-METHODS, PROBABILITY-DISTRIBUTIONS, ENERGY, PARAMETERS, GENERATION, PERFORMANCE, REGIONS
  • Karadeniz Teknik Üniversitesi Adresli: Hayır

Özet

Weibull distribution is widely used for modeling the wind speed data in literature. However, in real life, the wind speed data may not always be modeled by using the Weibull distribution. In other words, it may not represent all wind speed characteristic encountered in nature. Therefore, the alternative distributions are used in such cases. In this study, the Inverse Weibull (IW) distribution is used to model the wind speed. First, the modified maximum likelihood (MML) estimators of the parameters of IW distribution are obtained. Then the efficiencies of the MML estimators are compared with the well-known maximum likelihood (ML) and the least squares (LS) estimators via Monte-Carlo simulation study. Simulation results show that the LS estimators are the least efficient among the others. Finally, Weibull and IW distributions are used for modeling the seasonal wind speed data sets obtained from the Turkish State Meteorological Service. It is shown that IW distribution based on the ML and the MML estimates of the parameters provides better modeling than the Weibull distribution based on the corresponding estimates in most of the cases. The suitability of these distributions is compared with respect to root mean square error (RMSE) and coefficient of determination (R-2) criteria. (C) 2016 Elsevier Ltd. All rights reserved.