This paper describes some evidence of fractal order features in wind speed time series recorded at different observation stations both in USA and in Italy. Analysis were performed by using mono-fractal, multi-fractal and power spectra approaches. Results show that the average value of the box dimension for daily and hourly mean wind speed is D = 1.19 and D = 1.41 respectively, thus indicating that this kind of time series are fractal. The estimated average value of the Hurst exponent is H = 0.81 and H = 0.75 for daily and hourly time series respectively. From these Hurst exponents it is possible to infer the persistent behavior of wind speed. Furthermore, multi-fractal analysis shows that wind speed exhibits a bell-like shape spectrum with average width Δα = 0.47. Power spectra analysis has pointed out that wind speed time series behave as 1/f13 noise with average value of the β exponent of 0.46 and 1.37 for daily mean and hourly mean time series respectively. These latter results can be interpreted by saying that wind speed time series are Brown noise like

No abstract

No abstract

No abstract

Abstract-This paper deals with the problem of clustering daily wind speed time series based on two features referred to as Wr and H, representing a measure of the relative daily average wind speed and the Hurst exponent, respectively. Daily values of the pairs (Wr, H) are first classified by means of the fuzzy c-means unsupervised clustering algorithm and then results are used to train a supervised MLP neural network classifier. It is shown that associating to a true wind speed time series a time series of classes, allows performing some useful statistics. Further, the problem of predicting 1-step ahead the class of daily wind speed is addressed by introducing NAR sigmoidal neural models into the classification process. The performance of the prediction model is finally assessed.