Joint probability density forecast of short-term multiple wind farms output power

A method is proposed for forecasting joint probability density function (PDF) of short-term multiple wind farms' output power. Firstly, support vector machine is used to forecast single point value of the wind generation for each wind farm, and the PDF of prediction error is forecasted by sparse Bayesian learning; then the marginal PDF of wind generation is obtained. Secondly, the statistical property of prediction error of multiple wind farms' output power is analyzed to find the existence of temporal and spatial correlation properties. A dynamic conditional correlation regressive model is used to estimate the dynamic conditional correlation matrix, which can describe the quantitative relation of the temporal and spatial correlations. Finally, with the combination of PDF of each wind farm's output power and conditional correlation matrix, the joint PDF of multiple wind farms output power can be formed, and it is further transformed into multi-dimensional scenarios using multivariate random variable sampling method. Test results illustrate the efficiency of the method. © 2014 State Grid Electric Power Research Institute Press.