基于时间序列分析的风电功率预测模型

Wind Power Forecasting Model Based on Time Series Analysis

风电功率的精确预测对于电网合理调度,降低电网运行成本和保证电网系统安全性等方面有重要的意义。基于风电功率历史数据具有的时间序列特性,应用时间序列法建模,对风电场的发电功率进行短期和中长期预测,同时分析单一机组和汇聚机组对预测效果的影响。实证研究结果表明,利用时间序列分析方法,能比较精确地预测风电场发电功率,且短期内预测平均绝对百分比误差不超过15%,预测的合格率已超过85%,并且集中开发的方式可以减少预测误差。

Wind power reasonable accurate forecasting is an important part to reduce the power grid operation cost and ensure its safety in operation.It establishes a wind power prediction model based on time series analysis.Also the short-term and the mid-long-term wind farm power output are forecasted.Meanwhile,the error effect of four wind power units and fifty-eight wind turbines for wind power forecasting is analyzed respectively.The empirical research result shows that the shortterm and the mid-long-term wind power are more accurately forecasted with the method of time series analysis.The prediction of mean absolute percentage error is less than 15% in the short term and the prediction of qualified rate is more than 85%as well.It is easy to see that the prediction error of wind power will reduce in the condition of the concentrated development of the wind turbines.