基于支持向量机的发动机磨损故障识别

Wear Faults Recognition of Self-propelled Gun Engine Based on Support Vector Machine

针对发动机磨损故障识别问题,利用支持向量机(Support Vector Machine,SVM)的小样本训练学习的优势,建立基于SVM的发动机磨损故障识别方法,有效解决了困扰发动机油液分析故障诊断中小样本识别问题,为发动机故障诊断提供了一种新的途径。

In the wear fault diagnosis of engine, spectrometric oil analysis is so effective a diagnosis tool that we can obtain the concentration of about 20 elements. However, due to the limited number of the oil samples, it is difficult to build an effective mode recognition method. Then in this paper, we use SVM to set up a mode recognition method based on SVM and detect the wear faults in the engine. The results prove that the method can recognize the fault states from the normal states.