Data analytics methods for wind energy applications

In the wind industry, it is important to assess a turbine systems response under different wind profiles. For instance, a wind-to-power relationship is crucial for wind power forecast, and a wind-to-stress relationship is important for selecting critical design parameters meeting the reliability requirement. Given the complexity involved in a turbine system, it is impossible to write a neat, analytical expression to underlie the abovementioned relationships. Almost invariably does the wind industry resort to data driven methods for a solution, namely that wind data and the corresponding turbine response data (bending moments or power outputs) are used together to fit empirically the functional relationship of interest. This paper presents a couple of nonparametric data analytic methods relevant to wind energy applications with real life example for demonstration.