Correlation modeling among multi-wind farms based on Copula-ARMA wind speed model

In recent researches, wind speed model of single wind farm has been used to study the traditional wind farms modeling, ignoring the correlation between series of multiple wind speeds. Based on Copula function and ARMA model, a dependent model for multivariate wind speed series is proposed. Firstly, Copula multivariate distribution model is used to capture spatial dependencies, and then ARMA model is utilized to describe the temporal correlation. Finally, pick the spatial dependence of copula and the temporal characteristic of ARMA model to build Copula-ARMA wind speed model. The model is verified by the actual wind speed, results show that the proposed model can maintain the auto correlation, cross correlation and statistic characteristics of wind speed series. Research results can be applied to site selection, reliability assessment and capacity credit evaluation of large-scale clustered wind farms, providing basis for planning of large-scale wind power base.