Feature Extraction of Time Series Data for Wind Speed Power Generation

Feature extraction in a time series data, due to its sequential nature and missing data, is very essential for any decision process and control. Many workers have carried out this process, using various techniques, in the past, but a comparative study and analysis along with its application has been missing. The present work has carried out this process for wind speed time series data and has established a best suited type of wavelet transform for the data mentioned. Further this study shall be helpful in decision making for selection of a technique for feature extraction.