短期电力负荷的组合预测方法

电力负荷的影响因素很多,用单一预测模型进行预测时,因自身的局限,使其预测精度和稳定性不高。组合预测可以综合考虑各单一预测模型的结果,提高预测精度和稳定性。在电力预测中常用的指数平滑法、趋势分析法、Box-Jenkins时间序列法以及灰色预测法的单一预测基础上,建立组合预测模型,采用最小二乘法确定组合预测模型权重,并对某水电电力公司的上网负荷进行预测。实例表明,组合预测法能在短期电力预测中改善预测精度,提高预测的稳定性。

THE COMBINED FORECASTING METHOD FOR SHORT-TERM POWER LOAD

As there are many influence factors in power load, its forecasting precision and stability are usually poor if the forecasting is conducted only by the models with a single factor. The combined forecasting models can achieved better forecasting precision. Based on the methods of exponent smoothing, trend analysis, Box-Jenkins time-series and grey theory, an conbined model weighted by the least square method is built and applied to forecast the electric load of a water-power corporation. The example shows that the combined forecast can improve the forecasting accuracy and its stability in the electric load short-term forecasting.