Wind speed and direction predictions based on multidimensional support vector regression with data-dependent kernel

Abstract

The development of wind power has a higher requirement for the accurate prediction of wind. In this paper, a trustworthy and practical approach, Multidimensional Support Vector Regression (MSVR) with Data-Dependent Kernel(DDK), is proposed. In the prediction model, we applied the longitudinal component and lateral component of the wind speed, changed from original wind speed and direction, as the input of this model. Then the Data-Dependent kernel is instead of classic kernels. In order to prove this model, actual wind data from NCEP/NCAR is used to test. MSVR with DDK model has higher accuracy comparing with MSVR without DDK, single SVR, Neural Networks.