【摘要】 针对区间型数据的组合预测，该文引入了统计学中衡量两个概率密度函数之间相似程度的指标——Hellinger距离作为构建组合预测模型的最优化准则.在已构建的诱导有序加权连续区间的广义有序加权平均(IOWC-GOWA)算子的基础上，将区间数序列转化为实数序列，建立了一种新的区间数组合预测模型，并通过具体的数据进行分析，验证了该模型的有效性.结果表明：该文所建立的基于Hellinger距离和IOWC-GOWA算子的区间组合预测模型能在单项预测的基础上有效提升预测精度，是一种优性组合预测模型.更多还原

【Abstract】 For the combined prediction of interval data, this paper introduces Hellinger distance, an indicator in statistics to measure the similarity between two probability density functions, as the optimization criterion for constructing a combination prediction model, and converts the two-dimensional interval number sequence into a one-dimensional real number sequence on the basis of the constructed generalized ordered weighted continuous interval(IOWC-GOWA) operator, and establishes a new interval number combination prediction model, and conducts comparative analysis through actual data.The validity of the model is verified.The results show that compared with the single prediction method, the combined prediction model can effectively improve the prediction accuracy and is a superior combined prediction model.更多还原