A hybrid SVM-NARX based prediction method for Indian wind power sector. Issue 2 (17th February 2019)

• Record Type: Journal Article
• Authors: Kumar, Vijay; Pal, Yash; Tripathi, Madan Mohan
• Is Part Of: Journal of statistics & management systems. Volume 22:Issue 2(2019)
• Page Start: 363
• Page End: 378
• Publication Date: 2019-02-17
• Subjects: Forecasting -- Wind speed -- Neural Network -- Support Vector machine -- Optimization (SVM) -- MAPE -- Regression; Statistics -- Periodicals; Mathematical models -- Periodicals; Mathematical models; Statistics; Periodicals; 519.5
• DOI: 10.1080/09720510.2019.1580910
• Languages: English
• ISSNs: 0972-0510

Abstract

Abstract: Today energy demand in different sector of the world has been increased very much and supply problem become critical issue of energy supplying farms. Due to rapid decreasing of current available conventional sources and its harmful effect on human being and global warming, today people of the world pay more attention on pollution free sources of energy in term of nonconventional and sources of green energy. In this regard wind can be one of the cleanest and pollution free that will not generate any harmful emission and has some potential to reduce the dependence on polluted conventional sources. Although wind power generation faces main challenges in term of its unpredictable nature, frequency stability and availability in given time span. To overcome such challenges, the prediction of wind energy with certain accuracy is very essential. This work suggests a hybrid based method for prediction of wind power and its speed up remarkable certainty accuracy. This hybrid method considers the most useful data set from different available data set to train and validate of SVM-NARX model.