×

A nonparametric circular-linear multivariate regression model with its application to wind energy. (English) Zbl 1274.62297

Summary: A nonparametric circular-linear multivariate regression model using kernel weighted local linear method is proposed. The bandwidths are selected using the cross-validation method. The suitability of models is judged from the coefficient of determination \(R^2\). The model is applied in this paper to wind data extracted from National Centers for Environmental Prediction (NCEP) reanalysis data. The conclusion reached is that, compared with the existing model, the nonparametric model used in this paper provides a better fit and prediction for wind energy generated given wind speed, time and wind direction.

MSC:

62G08 Nonparametric regression and quantile regression
62H10 Multivariate distribution of statistics
62H11 Directional data; spatial statistics
91B76 Environmental economics (natural resource models, harvesting, pollution, etc.)
PDFBibTeX XMLCite
Full Text: Link