西安理工大学学报2017，Vol.33Issue(2)：138-144,7.DOI:10.19322/j.cnki.issn.1006-4710.2017.02.003
基于非负矩阵分解原理的方案优选方法及其应用
Method for optimal selection of schemes based on the non-negative matrix factorization principle and its applications
摘要
Abstract
The method based on the non-negative matrix factorization (NMF) principle is presented for the optimal selection of multi-index schemes and the development of optimization.The scheme-index matrix (SIM) is constructed after removing all the useless indexes from the original evaluation indexes.The optimal selection model is built based on the NMF principle,with the normalized SIM.Based on the cuckoo search algorithm,the normalized SIM is decomposed into a base vector and a weight vector,ensuring the minimum decomposition deviation of the model.Finally,the base vector is considered as the common unit of all the column vectors of the normalized SIM,and each element of the weight vector represents the corresponding scheme's superior degree.The NMF method is employed for the optimization of water resources engineering schemes,with the results showing that the evaluation processes are objective and fair,and that the evaluation results are reliable and convenient even without considering the weights of indexes.Therefore,the NFM method could be validated with other optimization selection methods,and could provide reliable decision results.关键词
节水灌溉/供水工程/水资源配置/方案优选/非负矩阵分解/布谷鸟算法Key words
water saving irrigation/water supply engineering/water resources allocation/optimal selection of schemes/non-negative matrix factorization/cuckoo search algorithm分类
自科综合引用本文复制引用
解阳阳,黄强,李向阳,刘赛艳,王义民..基于非负矩阵分解原理的方案优选方法及其应用[J].西安理工大学学报,2017,33(2):138-144,7.基金项目
国家自然科学基金资助项目(91325201) （91325201）
水利部公益性行业科研专项资助项目(201501058) （201501058）
陕西省重点科技创新团队研究方向为(2012KCT-10) （2012KCT-10）
陕西省教育厅重点实验室资助项目(13JS069) （13JS069）
