磨损微粒识别的模糊决策树方法

Approach of Wear Particle Classification by Fuzzy Decision Tree

磨损微粒识别是在磨损工况监控中确定机器磨损类型的关键步骤。本文基于非单调模糊推理方法,依据磨粒识别过程中所依赖特征的层次关系,建立了磨粒识别的模糊推理决策树.按照这一决策树开发的软件试验结果表明,本文所提出的方法在实现磨粒的自动识别方面是有效的。

Wear particle classification is a key to the identification of wear modes of a machine in wear condition monitoring. In this paper, the non-monotonic fuzzy reasoning method is discussed. Then a fuzzy decision tree applied to classify wear particles from used lubricant is presented according to the hierarchy of the attributes of wear particles for classification. It has been proven that the approach illustrated in this paper is effective for automatic wear particle classification by software tests.