Research On Short-term Wind Power Forecasting

As People’s continuous deep developing understanding of energy problems,clean energy has has attracted more and more attention from all countries.as clean energy,wind power has got extremely rapid development in the nearly more than ten years.Wind power brings great opportunities to people,at the same time,it also brought great challenge to power grid.Wind energy is a strong randomness renewable energy,People convert it into electricity and onto the electrical power grid has a strong impact on the grid.It can reduce its harm to power grid very well if there is a more accurate prediction of wind power.Therefore, how to grasp the law of the wind power effectively and reduce its harm to power grid has becomes the focus of the wind power development.Compared with foreign countries that the development of wind power are earlier,the construction of wind farms in our country is relatively late,so, the information such as wind speed, wind direction, temperature can hardly have good statistics,therefore, this article uses history wind power date to have short-term wind power prediction simulation.Wind power has strong randomness, the sudden change of weather can lead to rapid changes of the wind power, as for the feature of wind power,This article selects Hilbert-Huang transform which has better treatment effect to have a forecasting research on the history wind power data, The method can extract different components of history power data by EMD, then build different prediction models according to the characteristics of each component.The article proposes a combination forecasting model based on Hilbert-Huang transform to forecast wind power history data,a method is proposed in this paper in order to restrain the end effects of Empirical Mode Decomposion(EMD). The power sequence data is decomposed into seven intrinsic mode function components with different law characteristics by the empirical mode decomposition(EMD) method,by conscientiously grasping the law of each compoment, then different sequence can be forecasting by appropriate models and obtain the final prediction value of the next 72 hours by adding up the prediction results of each component.The model uses the actual data of wind farm in Henan to test. in order to comparison and analysis with the method adopt in this paper, chose two kinds of single forecasting model model to forecast the historical power data.The simulation results indicate that the short-term wind power forecasting model established in the paper has higher prediction accuracy and has Has certain rationality.