A combination forecast method based on cross entropy theory for wind power and application in power control

Predication of large-scale grid-connected wind power is important for a power system operation’s stability and security. To avoid the problem of being excessively dependent on reference samples, cross entropy theory-based information fusion technology is utilized to set up a new combination predication model for wind power prediction. In this new model, wind power prediction is regarded as an information fusion problem, and the weights of each predication method are dynamically determined by the cross degree of prediction methods obtained by cross entropy theory. The case studies of a practical wind farm validate the effectiveness and correctness of the proposed method to predict wind power and the design of the power dispatching scheme.