Prediction intervals or distribution functions of wind power in the future can be provided through probabilistic forecasting of wind power. Relevant research in China is still at the early stage. This paper gives a comprehensive review on the basic approaches

typical patterns and key problems in probabilistic forecasting of wind power. Firstly

the definition of probabilistic forecasting is presented and its applicable problems are summarized. Secondly

two different classification methods are introduced: conditional classification method and parametric classification method. New algorithms and evaluation indices used in probabilistic forecasting are also described. Finally

according to the-state-of-the-art of probabilistic forecasting

the shortcomings of error analysis and the insufficiency of probabilistic forecasting in combining with the power system

the future key issues and the research content which needs further exploration are summarized.