期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2008
卷号：8
期号：8
页码：162-169
出版社：International Journal of Computer Science and Network Security
摘要：Wind speed prediction from past observations has applications in many diverse fields such as Target tracking, Missile guidance, Satellite launch, Air traffic control, Weather forecasting, Ship navigation, Electrical power generation using wind energy. The wind speed is determined by many other atmospheric variables, such as pressure, moisture content, humidity, rainfall etc. Literature reports a number of models for wind speed prediction. In this paper a comparison is carried out between Neural network models and Support Vector Machine models built for predicting wind speed in short term. Analysis shows that SVM models compute faster & give better accuracies than the Neural Network models.
关键词：Short term wind speed prediction, Neural networks, Back propagation, Support Vector Machine [SVM], forecasting, hyper plane, kernels, classification
