SVM Tuned NARX Method for Wind speed & power Prediction in Electricity Generation

Due to continuous depleting of conventional energy reserves as well as global warming issues has diverted world attention towards non conventional energy sources. Out of different non conventional energy sources wind can be consider the cleanest source with minimum possible pollution or harmful emissions and has the potential to decrease the relying on conventional energy sources. Wind energy can play a major role to meet our energy demands, however, it faces various problems such as intermittent nature, instability in frequency. To reduce such issues the idea of futuristic wind speed trends and weather conditions are required. This paper suggests a method for wind speed, power forecasting using NARX and Support Vector Machine (SVM). This hybrid technique can select most suitable data segment from available data. These data segments are used for the training and validation of SVM model. The historical data used is from Kolkata region wind energy farm.