Research On The Application Of Statistical Methods For Wind Farm Power Prediction

With the development of wind power generation and faster,the share in the grid is growing, but the intermittency and randomness of wind power makes its security on power system running and scheduling big challenge. China has abundant wind energy resources, the development of wind power industry has a good resource base, not only the average wind speed is high, a wind power density is large, the development and utilization of wind power also has the very big development space. In addition, for a long time of single energy structure, new energy and traditional energy development lag and such problems as environmental pollution, has seriously hindered the economic and social sustainable development.In this article, through an in-depth study on characteristics of wind power,completed the following several aspects: first, the effects of weather conditions on the predictive power generated by the wind farm is proposed to establish four wind farm power prediction model, wherein Nonlinear method is the most ideal. The second part introduces a variety of wind power prediction technology and methods, and focuses on the statistical method of SVM and ANN methods. Third, the design is based on the statistical method of wind power prediction system, which forecast system is introduced in detail the module design, combined with Support Vector Machine(Support Vector Machine, SVM) algorithm and Neural Network(Neural Network) to predict the wind power, finally analyze the model and the prediction effect.Based on the wind farm power output prediction as the research object, put forward the wind power prediction method based on statistical method, through analyzing test data, more significant results have been achieved.