MULTI-SCALE PERIODICITY RECOGNITION METHOD OF WIND SPEED BASED ON MORLET WAVELET WITH TIME-FREQUENCY FACTORS

The volatility of wind speed poses great challenges to the planning and operation of the power system with new energy as the main component. The periodicity of wind speed is an important rule to characterize its fluctuation. However, there are few studies on the periodicity of wind speed, and have no quantitative index to evaluate it at present. To solve the above problems, two quantitative evaluation indexes, periodic intensity (PI) and relative periodic intensity (RPI), are proposed in this paper based on the Morlet wavelet. Furthermore, a time-frequency factor is added in the mother wavelet in order to adapt the periodicity analysis of wind speed which with high variability. The wind speed data of south Gobi, Mongolia is taken as examples to analyse the periodicity of wind speed at multiple time scales, including year, season and month. Then, the wind speed prediction accuracy in different time windows, traditional year time window and the calculated periodic time window, are compared to verify the validity and superiority of the proposed periodicity evaluation indexes. The results show that, the prediction accuracy is improved by 20.51% and 15.79% at two wind measurement points by using the periodic time window.