基于K-L信息距离的证据冲突特征识别与应用  

Identification and application of the evidence conflict based on K-L information distance

作　　者：李杨[1,2] 郭亚军[1] 杨友良[2] 

机构地区：[1]东北大学工商管理学院,沈阳110004 [2]河北联合大学,唐山063009

出　　处：《系统工程理论与实践》2014年第8期2071-2077,共7页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(71071030);河北省科技支撑计划项目(12276906D);河北省教育厅资助科研项目(QN2014028)

摘　　要：证据冲突是多证据融合中各种合成方法的主要制约问题,本文引入K-L信息距离函数描述证据间的冲突特征,通过构造距离矩阵定义独立证据在整个系统的冲突程度,实现对证据冲突的系统化识别.模拟实验表明:K-L信息距离的有效冲突识别,能完善D-S理论合成规则的应用约束,对正常冲突证据合成得到优化收敛结果,对高度冲突隔离证据的特征分析提取价值信息.Evidence conflict is proved to be the main restricting problem of various synthesis methods in the evidences integration, this paper introduces the K-L information distance function to describe the conflict characteristics between evidences, defines the conflict degree of independent evidence in the whole system through the structure of distance matrix, realizes the systematic recognition of evidence conflict. Simulating results show that： the effective conflict identification of K-L information distance can complete the application constraint of D-S theory＇s synthetic rule, obtaining the optimized convergence results for synthesis of normal conflicting evidences, extracting the valuable information for the characteristics analysis of highly conflicting isolation evidence.

关 键 词：D—S证据融合 信息距离 冲突识别 隔离 

分 类 号：C934[经济管理—管理学]