基于信息融合的柴油机故障诊断技术

Diesel Fault Diagnosis Based on Information Fusion

针对传统柴油机故障诊断方法诊断准确率不高的问题,提出了一种多传感器信息融合的方法,将模糊神经网络与D-S证据理论结合起来,使两者优势互补；通过简化网络结构提高局部诊断网络的诊断能力,并使证据理论的基本可信度分配不再完全依赖专家进行的主观化赋值,实现赋值的客观化；通过建立的多传感器信息融合系统对柴油机的故障进行识别,使不同的信息源相互补充,很好地建立故障与征兆之间的定量映射关系,获得对故障状态的最优估计和判决。

For the reasons of low accuracy of traditional diesel fault diagnosis, a multi--sensor information fusion method was presented , which combined FNN with evidence theory using their superiority and avoiding their disadvantages. The local diagnosis networks was advanced by simplifying network structure and realizing evaluation objectification by the basic reliability distribution of the evidence theory not completely depending on the expert subjective valuations. This method is applied to Diesel Fault Diagnosis, which taking full advantages of various information can establish quota mapping relations between the diagnosis and the indication and obtains the most superior estimate and the decision.