基于已有的轧钢加热炉的能耗数据,引入了各预测方法的预测相对误差、预测对象的变化趋势和灰色基本权重等概念,建立了轧钢加热炉的能耗模糊变权重组合预测模型。结果表明,此组合预测平均误差和预测平方根误差均较小。 Based on the energy consumption data of existing rolling reheating furnaces, the concepts of forecasting relative error, forecasting trend and gray basic weight of each forecasting method are introduced, and the fuzzy variable weight combination forecasting model of rolling reheating furnace is established. The results show that both the average forecast error and the square root of prediction error of this combination are small.