Z网络下新的推理算法在不确定决策中的应用  

Application of New Reasoning Algorithm in Uncertain Decision Making Based on Z-network

作　　者：张最 吴涛[1,2] 涂成凤 ZHANG Zui;WU Tao;TU Chengfeng(School of Mathematical Sciences,Anhui University,Hefei 230601,China;Key Laboratory of Intelligent Computing and Signal Processing,Ministry of Education,Anhui University,Hefei 230039,China)

机构地区：[1]安徽大学数学科学学院,安徽合肥230039 [2]安徽大学计算机智能与信号处理教育部重点实验室,安徽合肥230039

出　　处：《佳木斯大学学报（自然科学版）》2021年第5期52-56,79,共6页Journal of Jiamusi University：Natural Science Edition

基　　金：国家自然基金面上项目(71871001);国家自然基金青年项目(618006001);省级质量工程项目2019jyxm0068。

摘　　要：针对Z-number的模糊不确定性和概率不确定性推理,在Z网络模型的基础上对其推理算法上做出了一些改进。首先,在Z-number理论基础上对离散Z-number在if-then规则下进行最大熵方法处理的算法过程做出了一些改进,由于算法过程中优化模型获取的区间值概率随着给定数据的变化并不一定能满足最大熵方法中的约束条件,所以对这些区间值概率进行了优化。其次,基于Z网络的结构和概率推理过程,对利用Z网络获取区间值概率然后应用最大熵方法获得约束部分最可能的潜在概率分布过程作出了同样的改进。最后利用一个实例来说明改进后的方法的有效性和可行性。Aiming at the fuzzy uncertainty and probabilistic uncertainty reasoning of Z-number,some improvements are made on the reasoning algorithm based on Z-network model.Firstly,on the basis of Z-number theory,some improvements are made to the algorithm process of maximum entropy method for discrete Z-number under if then rule.Because the interval value probability obtained by the optimization model in the algorithm process does not necessarily meet the constraints of maximum entropy method with the change of given data,these interval value probabilities are optimized.Secondly,based on the structure and probability reasoning process of Z-network,the process of obtaining interval valued probability by Z-network and then obtaining the most likely potential probability distribution of constraint part by maximum entropy method is improved.Finally,an example is given to illustrate the effectiveness and feasibility of the improved method.

关 键 词：Z-number Z网络 潜在概率 Z-valuation不确定性推理 

分 类 号：C934[经济管理—管理学]