Bispectrum and Energy Analysis of Wind Speed Data

Abstract

The bispectrum has a different analysis functions compared to the power spectrum analysis. The bispectrum provides additional information for nonlinear or nonGaussian data, and also the method gives detailed information on the phase information signal. In this study, bispectrum and energy analysis was performed for wind speed data for the regions of Turkey Kırklareli. Hourly wind speed data of Kırklareli Region belonging to the year 2011 is used. Data were obtained from The General Director of Meteorology, Ankara, Turkey. As well as time-amplitude graph, energy graph obtained and monthly analysis is made of this diagram. Subsequently, sufficiency level of speed zones has been tested frequency from time and frequency information by obtaining spectrograms of velocity information. Random processes which are described more accurately than statistically and high grade statistics for processing phase information (more than two) and studies on spectrum can provide useful results. In this context, bispectrum of annual velocity information was obtained. In addition to the bispectrum analysis, methodology of the diagonal slices spectrum is used for data which do not show mark in the non-linear structure or Gaussian distribution researches. Also, after obtaining bispectrum surfaces, monthly results through these charts were found and evaluated by removing diagonal slices analysis. Peaks in the bispectrum indicates the frequency component and phase overlap in the signal. By means of these analyses, the frequency component and phase conflict are determined in the wind speed data. The results of the analysis determined by the energy level was consistent with the wind speed. Wind speed phase-coupling and diagonal slices are calculated. As a result of the calculations, it is determined low and high frequency groups.