Wind Power Forecasting Model Based On Time Series Analysis

In recent years, with the rapid development of energy technology in our country, wind power as a kind of clean and renewable energy attaches great importance. It becomes one of the large-scale development and utilization of the objects in a short time, since it is rich in resources and it need not be mined and transported. Today, wind power is the fastest growing and most mature of renewable energy power generation technology, compared with conventional power, not only its no polluting is favored by people, but also the cost is low that it is available to provide the opportunity of development and utilization on a large scale. However, Wind power has its own disadvantages, when wind power output, since it has the characteristics of volatility and peak shaving, and these features make it hard for the wind farm be outputted wind power steadily. This gives the related operation department of power system many difficulties in planning wind power. The work of state grid dispatching department has been a huge challenge when large scale wind power access to the power system. Therefore, it is significant for the scholars to strengthen the precise prediction research of wind power, because it can schedule the wind power system of power generation planning, reduce the wind power unit spare capacity and running cost and guarantee the safe and stable operation of power system.First of all, under the background of the national wind power with a large-scale development, the paper is launched with the investigation and study of Shanxi power grid, we realized that it is of great significance for the safety of Shanxi power grid as well as the national power grid to accurate wind power prediction. Secondly, standing in the angle of the existing research results for wind power prediction at home and abroad, we carried on the literature review of wind power prediction. Then we carried out the wind power prediction method and the numerical weather prediction for the influence of wind power prediction. Thirdly, based on the historical data of the wind power, which is with the time sequence features, it applied the method of time series in modeling for the wind power to forecast in short-term and long-term projections. It also analyses the effect between a single unit and the unit’s influence on the prediction at the same time. Then, the result of the empirical study shows that with the method of time series analysis, it can more accurately predict the wind power. The mean absolute percentage error is less than 15% and the percent of pass is also more than 85% in the short term prediction. We knew the concentrated development way can reduce the prediction error. Then, it analyses the existed deviation based on the model prediction results and the actual real value to explore the reasons of the existence of the prediction deviation. Finally, it completed the design and implementation of the wind power prediction system based on the GUI interactive and gave the graphical user interface based on intelligence. It also provided technical support and personalized service for the wind power system management staff to forecast.