基于模糊综合函数的故障诊断

在决策级信息融合的故障诊断中,根据各传感器所作决策不可避免地带有主观性和模糊性。针对这种情况,如何准确地判断是否存在故障并尽快找到故障源就取决于合适的融合方法。提出了一种新的属于决策级的信息融合方法,即:先求出各传感器所作决策的可能性分布,然后利用模糊综合函数来进行融合故障识别与诊断,从而可解决多信息源信息融合中普遍关心的信息可信度以及融合准则问题。

Fault diagnosis based on fuzzy integrated function

In fault diagnosis based on decision-level information fusion, sensor-based decision always has subjectivity and ambiquity. A new decision-level fusion method is put forward: After probability distribution of each sensor is decision is obtained, a fuzzy integrated function is used to fuse the information and diagnose the fault. In this way, the problems of how to decide the information confidence level and how to select fusion rule in multisensor information fusion can be solved.