Estimating wind speed probability distribution using kernel density method

Accurate estimation of long term wind speed probability distribution is a fundamental and challenging task in wind energy planning. This paper proposes a nonparametric kernel density estimation method for wind speed probability distribution. The proposed method is compared with ten conventional parametric distribution models for wind speed that have been presented in literatures so far. The results demonstrate that the proposed non-parametric estimation is more accurate and has better adaptability than any conventional parametric distribution for wind speed.

► This paper proposes a nonparametric kernel estimation for wind speed probability distribution.
► This paper develops a new technique to determine an optimal bandwidth based on wind speed data.
► The proposed method is compared with ten conventional parametric distribution models for wind speed.