首页 | 本学科首页   官方微博 | 高级检索  
     

证据推理理论及其方法研究
引用本文:金宏斌,蓝江桥,李鸿飞,赵国红. 证据推理理论及其方法研究[J]. 空军雷达学院学报, 2011, 25(3): 167-171,179. DOI: 10.3969/j.issn.1673-8691.2011.03.004
作者姓名:金宏斌  蓝江桥  李鸿飞  赵国红
作者单位:1. 空军雷达学院四系,武汉,430019
2. 空军雷达学院,武汉,430019
3. 94981部队,南昌,330200
摘    要:在实际融合系统中,往往需要处理各种不确定信息,而证据推理理论为不确定信息的表达和组合提供了自然且强有力的方法,得到了广泛的应用.为此,研究了目前常用的两种证据推理理论(DS理论和DSm理论)及其方法.基于这两种理论原理,从修改组合规则和修改证据模型两种思路,阐述了基于DS理论的常见改进方法;从理论研究和工程应用两方面给出了DSm理论的研究现状.最后,归纳了证据推理理论的不足,展望了其未来发展方向.

关 键 词:DS理论  DSm理论  组合规则  证据冲突

Study on Evidential Reasoning Theories and Their Methods
JIN Hong-bin,LAN Jiang-qiao,LI Hong-fei,ZHAO Guo-hong. Study on Evidential Reasoning Theories and Their Methods[J]. Journal of Air Force Radar Academy, 2011, 25(3): 167-171,179. DOI: 10.3969/j.issn.1673-8691.2011.03.004
Authors:JIN Hong-bin  LAN Jiang-qiao  LI Hong-fei  ZHAO Guo-hong
Affiliation:JIN Hong-bin1,LAN Jiang-qiao2,LI Hong-fei1,ZHAO Guo-hong3(1.No.4 Department,AFRA,Wuhan 430019,China,2.AFRA,3.94981 Unit of the PLA,Nanchang 330200,China)
Abstract:In the real fusion system,there exists frequently various uncertain information that is required for processing,while the evidential reasoning theories have offered the natural and powerful methods for the expression and combination of the uncertain information,and have been widely used.Therefore,this paper investigates the commonly used two types of evidential reasoning theories,such as Dempster-Shafer(DS) theory and Dezert-Smarandache(DSm) theory,and their methods.Based on the two theories’principles,the common improved DS-based theory methods are expounded in terms of two ways of modifying the combination rules and evidential models,and the research status of DSm theory presented from the two aspects of theoretical research and engineering application.Finally,the drawback of evidential reasoning theories is summed up and the future developmental trends expected in this paper.
Keywords:DS theory(DST)  DSm theory(DSmT)  rules of combination  evidential conflict  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号