在不确定环境组合预测中,用模糊权重系数更能体现各单项预测方法的客观表现。文章提出一种新的权重系数为三角模糊数的组合预测方法。首先建立以组合预测精确度指数最小为准则的模糊加权组合预测模型,为了避免样本数据中极端值对模型的影响,对模型进行改进,提出带有0-1变量的模糊加权组合预测模型。进一步考虑到单项预测方法在不同时刻的表现有所差异,建立基于诱导有序模糊加权平均(IOFWA)算子的模糊变权组合预测模型,该模型不仅能克服极端值的影响,而且具有更高的预测精确度。并实证验证了该方法的适用性和灵活性。
In uncertain environment combination forecasting, the fuzzy weighted coefficient better reflects the objective per-tormance of each single forecast method. This paper proposes a new combination forecast method with the weighted coefficients as triangular fuzzy numbers. Firstly the paper constructs a fuzzy weighed combination forecasting model by minimizing the combination forecasting accuracy index. In order to avoid the influence of extreme values on the model in the sampled data, the paper presents an ameliorated fuzzy weighed combination forecasting model with 0-1 variables. Moreover, in consideration of the different performance of each single foreeast method at different times, the paper proposes a fuzzy variable weight combination forecasting model based on the induced ordered fuzzy weighted averaging (IOFWA) operator, and this model not only overcomes the influence of extremum, but also has better precision. Finally an empirical study is done to validate the applicability and flexibility of the proposed method.