Wind Power Time Series Generation Based On Wind Power Characteristics And Copula

As the development of global economy, the non-regenerated energy has been gradually consumed, to solve the problem, new energy has been hot research field. Wind energy is clean and renewable, but wind power has brought great impact to power system due to the randomness, while the wind resource distribution is uneven in China, large wind power production bases are away from the center of the load position, causes problems of transfer, wind energy curtailment rates is high. To solve the problem, characteristics of wind power output must be analyzed, but the mostly research now is qualitative analysis, the result can’t contain wind power characteristics comprehensively, related work is needed to carry out. Main methods of generating time series are Markov chain Monte Carlo and ARMA(Autoregressive Moving Average), the disadvantage of the methods are obvious and the wind power characteristics can’t be describe comprehensively.In light of the above two problems, this paper mainly defines and analyzes characteristics of wind power data in 24 wind farms, which include mean value, standard deviation, a utocorrelation Function, wind power probability density function and wind power fluctuation, compare the wind power of different years. Come up with the method of wind power time series generation based on autocorrelation function, wind power probability density function and wind power fluctuation. Carry on simulation, compare the result with original data and data of Markov chain Monte Carlo method. The method is accurate and reflect the wind power characteristics very well.