基于向量平均投影和IOWA算子的区间数组合预测模型  

The Interval Number Combination Prediction ModelBased on Vector Mean Projection and IOWA Operator

作　　者：胡凌云[1] 蒋宗华[1] 张林[1] HU Lingyun;JIANG Zonghua;ZHANG Lin(School of Management Science and Engineering,Anhui University of Finance and Economics,Bengbu Anhui 233030,China)

机构地区：[1]安徽财经大学管理科学与工程学院,安徽蚌埠233030

出　　处：《鞍山师范学院学报》2025年第2期17-24,共8页Journal of Anshan Normal University

基　　金：安徽省哲学社会科学规划项目(AHSKY2020D42);安徽省高校自然科学重点项目(2022AH050602);安徽财经大学科研基金资助重点项目(ACKYB23011).

摘　　要：区间数是一类比较简单的模糊信息,构建区间数组合预测模型,可以在模糊预测问题中获得更符合实际的预测结果.通过将区间数的左右端点分开,引入向量平均投影和IOWA算子,针对实际值区间数的左端点序列和各单项预测值区间数左端点序列、实际值区间数的右端点序列和各单项预测值区间数右端点序列,分别构建向量平均投影和IOWA算子的多目标最优组合预测模型,利用偏好系数将多目标最优模型转换成单目标最优模型.通过实例数据加以验证,结果显示所构建的组合预测模型是合理有效的.Interval numbers represent a relatively simple form of fuzzy information.Constructing a combined interval number prediction model can yield more realistic prediction outcomes in fuzzy forecasting problems.By separately handling the left and right endpoints of interval numbers and incorporating vector average projection and induced ordered weighted averaging IOWA operators,we develop multi-objective optimal combination prediction models for both the sequence of left endpoints and the sequence of right endpoints from actual value intervals and individual predicted value intervals.Using preference coefficients,we transform the multi-objective optimal model into a single-objective optimal prediction model.Finally,empirical data demonstrates that the constructed combination prediction model is reasonable and effective.

关 键 词：组合预测 区间数 向量平均投影 IOWA算子 

分 类 号：O224[理学—运筹学与控制论] F224[理学—数学]