发动机运行状态的主分量灰色诊断模型研究

Research on PCA based Grey fault diagnose Model for Reorganization of Running State of Engine

针对灰色理论在故障诊断领域应用中存在的准确度较差问题,提出了一种主分量空间下的灰色故障诊断模型。该诊断模型的应用首先将高维相关特征参数转化为低维相互独立的特征参数,在此基础上利用灰色理论中的灰色关联方法进行故障模式的诊断。用该方法对某电喷发动机的运行状态进行了识别。实验结果表明,基于主分量分析方法的灰色故障诊断算法可以有效消除特征参数之间的相关性,找出故障描述的最佳特征参数,突出不同参数的重要性,从而提高灰色诊断结果的准确率,可以在实践中发挥有效作用。

Pointing at low accurate existed in grey theory based fault diagnose method, a PCA based grey fault diagnose method is devel- oped. In this method, multi--dimensional correlated characteristics parameter are transformed into low dimensional independent characteristics parameter firstly, and fault pattern is judged based on grey relational analysis. The running state of an electronic rejection engine is rec- ognized with this method, experiment result indicates that PCA based grey fault diagnose method can dispel the relationship existed in charac teristics parameters, find the optimization characteristics parameter for fault description, emphasize the importance of the characteristics parameters, improve the accurate of diagnose algorithm,and can play significant performance in actual application.