Assessment of Power Curve Fitting Performance of Parametric Models for Wind Turbines

In wind energy systems, wind speed variability and wind power fluctuation negatively affect the power system reliability. To overcome this challenge, actual wind turbine power curves serve as one of the important tools for condition monitoring and troubleshooting, easier forecast of wind power production and ensuring the stable operation of wind turbines. Motivated by this, this study compares the goodness-of-fit results of polynomial, Fourier, Gaussian and sum of sines parametric models in wind turbine power curve fitting. According to the accuracy results obtained, 9 th -degree polynomial, 8-term Fourier, 4-term Gaussian and 5-term sum of sines models show good parametric modeling performance in their own curve fitting category. Among them, 8-term Fourier model stands out by achieving the least power curve fitting errors. In addition, traditional benchmark models have been outdone in terms of the goodness-of-fit statistics.