基于改进的无量纲指标信息融合故障诊断方法

针对无量纲指标信息的不确定会产生冲突证据的问题,提出一种基于改进证据理论的无量纲指标信息融合故障诊断方法.采用人工免疫算法构建原始证据;通过计算无量纲免疫检测器证据之间的相似度找出相似证据和冲突证据.保留相似证据,并利用可信度修正冲突证据,从而降低冲突证据对证据理论合成结果的影响.组合式旋转机械故障诊断试验验证本文方法有效.该方法能够明显提高故障诊断准确率.

Fault diagnosis method based on improved dimensionless parameters information fusion

Aiming at conflict evidence resulting from uncertainty of dimensionless parameters, an information fusion fault diagnosis approach based on improved evidence theory was proposed. Firstly, using the artificial immune algorithm to create original evidence was presented. Then, conflict evidence and similar evidence were distinguished through calculating the similar degree of evidence. And the conflict evidence was modified by credibility, which can decrease the effect of composed result from uncertainly. Finally, simulation results show that the proposed approach is valid. The diagnostic accuracy was significantly improved.