Search

Abstract: The fluctuation of wind speed within a specific time period affects a lot the energy conversion rate of wind turbine. In this paper, the concept of fractal dimension in chaos theory is applied to investigate wind speed characterizations; numerical algorithms for the calculation of the fractal dimension are presented graphically. Wind data selected is observed at three wind farms experiencing different climatic conditions from 2006 to 2008 in Taiwan, where wind speed distribution can be properly classified to high wind season from October to March and low wind season from April to September. The variations of fractal dimensions among different wind farms are analyzed from the viewpoint of climatic conditions. The results show that the wind speeds studied are characterized by medium to high values of fractal dimension; the annual dimension values lie between 1.61 and 1.66. Because of monsoon factor, the fluctuation of wind speed during high wind months is not as significant as that during low wind months; the value of fractal dimension reveals negative correlation with that of mean wind speed, irrespective of wind farm considered. For a location where the wind distribution is well described by Weibull function, its fractal dimension is not necessarily lower. These findings are useful to wind analysis. [Copyright &y& Elsevier]

In this paper, 10-year meteorological data observed at Taipei, Taiwan, from 2006 to 2015, are studied to investigate the inherent correlation between renewable resources. The data are investigated in the time-frequency space using wavelet analysis including wind speed, global solar radiation and air temperature. Through the wavelet transform of the time series, the related cross wavelet transform, wavelet coherence and phase angle are obtained. The results show that the wavelet power spectra exhibit prominent oscillations at annual and half-year periods for the three kinds of meteorological data, while some oscillations for other periods do not show a regular pattern. Wind speed spectrum is more irregular than the other two types of data due to climatic and geographic factors, whereas temperature has the most regular form. From the cross wavelet transform of data pairs, it is seen wind speed and solar radiation have an anti-phase correlation, implying that wind and solar energy can be used in complementary roles in electricity generation in this region. The results of wavelet analysis coincide with the variation trend of the monthly mean of the meteorological data over the 10-year period. Wavelet analysis is a robust tool for the oscillation characteristic study on time series data. [ABSTRACT FROM AUTHOR]