Fitting Weibull distribution on wind speed data series
Akylas Evangelos, Zavros Panagiotis, Skarlatos Dimitrios and Marios Fyrillas
CFE-ERCIM 2010, held 10-12 December, 2010 in London, UK, abstract E756
This study is concerned with the two-parameter Weibull distribution which is widely employed as a model for wind speed data. The the maximum likelihood method for the estimation of the Weibull parameters was proved by numerous studies to be a more accurate technique than the commonly used log-linear regression. However it depends on detailed original information and demands more computational time. We present a modified technique which is based on the estimation of higher moments of the Weibull distribution. The technique demands only basic averaged statistical information, it is computationally cheap and renders practically equivalent results with the maximum likelihood method.
Akylas Evangelos, Zavros Panagiotis, Skarlatos Dimitrios and Marios Fyrillas
CFE-ERCIM 2010, held 10-12 December, 2010 in London, UK, abstract E756
This study is concerned with the two-parameter Weibull distribution which is widely employed as a model for wind speed data. The the maximum likelihood method for the estimation of the Weibull parameters was proved by numerous studies to be a more accurate technique than the commonly used log-linear regression. However it depends on detailed original information and demands more computational time. We present a modified technique which is based on the estimation of higher moments of the Weibull distribution. The technique demands only basic averaged statistical information, it is computationally cheap and renders practically equivalent results with the maximum likelihood method.