Analysis of Wind Farm Spatial Correlation Output Model Based on Copula Kernel Function

This paper proposes to use Copula kernel functions instead of empirical Copula to select models that describe the correlation between wind farm output. Taking the actual output data of two wind farms in North China as samples for calculation, the results show that the minimum norm between the Copula kernel function and the theoretical Copula function is smaller than the empirical Copula method, indicating that the optimal Copula selected based on the Copula kernel function can more accurately reflect the spatial correlation characteristics of the actual output samples of the wind farm; Compared with the minimum norm of a single Copula, the minimum norm of a mixed Copula is smaller, indicating that mixed Copula is more suitable for capturing asymmetric tail correlation structures. Finally, statistical tests were conducted on the selected optimal Copula, and the results showed that the mixed Copula selected using the Copula kernel function method had a higher degree of fit.