The method of forecast combining could make greatly use of the information of various samples, effectively reduce the influences of the environmental factors during the forming of the single forecasting model, and improve the precision of forecast combining. By studying the methods of Grey Model (1,1), recurrence of new information with equal dimension and regression analysis, firstly, GM(1,1) based on the original progression was built to forecast the short-term data, and then, by means of the recurrence of new information with e-qual dimension, the short-term data was added to the original progression to generate a new progression; lastly, a regression model based on a new progression was built to predict the long-term power load. Numerical studies reveal that the precision of the forecast combining approach is better than that with the grey model.%组合预测法能较大限度地利用各种预测样本信息,有效地减少单个预测模型建立过程中一些环境随机因素的影响,提高预测精度.通过对灰色预测法、等维新息法、回归分析法的研究,先由原始数列建立灰色预测模型,预测出近期数据,再运用等维新息思想,把灰色模型的近期预测值添加到原始数列中,生成组合的数列,由新生成的组合数列建立回归模型,预测长期电力负荷值.通过一个实例比较组合预测和灰色模型的预测结果,对比得知,组合预测的精确度优于灰色预测.
展开▼
