A NEW PROBABILITY DENSITY FUNCTION (PDF) MODELLING METHOD FORWIND SPEED

摘要

Defining a wind speed probability density function (PDF) model is a compact method ofproviding a scientific judgement of a regime resource and profile. Actual wind speeddistributions are highly heterogeneous and deformed in nature and, modelling this complexphenomena demands flexibility in the chosen models. It is found that the traditional PDFmodelling method of hourly mean wind speed (HMWS) observations on an annual basis isstatistically invalid in that the assumptions of randomness, homogeneity, independence andstationarity are severely violated. A suggested PDF modelling method of year-year hourlysamples within any particular month is proven to be valid and is recommended. A ten yearunpublished HWMS data set for February logged at the Crown Point meteorological station isused to illustrate and test the new method. The concept of a pseudo time series of repeatable PDFmodels is introduced for new practical applications in the field.