Statistical Analysis of Wind Turbine Operational Data

This paper deals with statistical data analysis for wind turbine. Conversion of wind power into electricity is a promising area of clean energy. Reliability and efficiency of wind turbines are provided by the operation system. Decision-making in these systems is based on the application of algorithms for empirical data processing. These algorithms are for detecting, parameter estimation, simulation, prediction and analysing mathematical models, etc. This paper considers the problem of making and choosing the best mathematical model for describing the diagnostic variable of wind turbine based on the approximation of empirical data with single and two-segmented regression usage. The two-segmented approximation is based on usage of the Heaviside step function for obtaining a general mathematical equation and a new method for determining the optimal switching point abscissa. Evaluation of unknown coefficients of approximation function was carried out on the basis of ordinary least-squares method.