基于证据理论数据融合的故障诊断研究

Study on Data Fusion Fault Diagnosis Based on Dempster-Shafer Evidential Theory

从D-S证据理论的基本概念和融合推理方法出发,构造了符合故障诊断特点的基本可信度分配,建立了多故障特征信息融合诊断框架,并将其应用于液压泵故障诊断,实例表明该方法能有效提高诊断可信度,充分显示了该诊断方法的有效性。

Starting with the basic conception and fusion reasoning method of Dempster-Shafer evidential theory. It is proposed to a basic probability assignment method of according with fault diagnosis characteristic. Multi-fault characteristic information fusion diagnosis frame is constructed and applied in the faih diagnosis of the hydraulic pump. It is shown that the method is effectively to improve the reliability of the diagnosis. So the fault diagnosis method proves highly effective.