存在冲突证据下的证据推理多属性决策方法

A New Evidential Reasoning Approach for Multiple Attribute Decision Making under Circumstance of Existing Conflict Information

针对传统的证据推理方法对证据冲突处理能力的不足,在引入冲突参数的基础上提出了新的证据推理算法,通过证明,新算法完全满足证据合成的4个公理。基于方案集间冲突参数对决策结果影响差异最小化原则,提出了新的冲突参数优化模型,最后通过实例仿真,论证了新算法对解决证据间存在冲突时的多属性决策问题的可行性和优越性。

The conflict between evidences can not be calculated in reason during the course ot multiple atmDute decision making（MADM） with conventional evidential reasoning（ER） approach. To overcome this shortcoming, a new algorithm is proposed based on introducing the conflict parameter. The new algorithm is proved to satisfy four synthesis axioms completely. To assure the reliability of the decision results, the effect of conflict parameter on the decision results should be minimum value, based on this principle, an optimize model which can calcu- late the conflict parameter is proposed. Two numerical examples illustrate the validity and superiority of the new method on dealing with the conflict evidences.