主观证据交互式提取及融合方法

• 作 者: 杜元伟 孙永河 段万春
• 机 构: 昆明理工大学管理与经济学院,
• 出 处: 控制与决策
• 发表日期: 2011年05期
• 关键词: 融合方法 证据理论 动态知识 证据信息 矩阵 动态知识矩阵 组合规则 识别框架 有效性 函数

为解决现有研究并未考虑客观证据源因受外界干扰而无法获取有效证据信息的问题,基于动态知识矩阵构建了能够从决策专家知识、经验、直觉中逐步获得主观证据信息的提取方法,并给出了由动态知识矩阵向基本概率分配函数转化的推断定理,在此基础上,通过设置冲突因子阈值构造了能够平衡推断成本与推断效果之间矛盾关系的主观证据交互式融合方法.数值模拟结果表明了所提出方法的有效性和可行性.

In order to solve the problem that the existing research does not consider that the objective evidence source can not obtain the valid evidence information because of the outside interference, this paper constructs the extraction method which can obtain the subjective evidence information gradually from the knowledge, experience and intuition of decision experts based on the dynamic knowledge matrix The deduction theorem of conversion from dynamic knowledge matrix to basic probability distribution function is given. Based on this, an interactive fusion method of subjective evidence can be constructed by setting the conflict factor threshold to balance the contradiction between cost and inference effect. The results show the effectiveness and feasibility of the proposed method.