【摘要】 引入了一套磨粒的形态学参数描述体系 ,实现了磨损颗粒图象的数值描述 ,并将灰色系统中的定权聚类技术用于磨损颗粒的自动识别 ,编制了相应的计算机模拟程序。在识别过程中根据某型航空发动机的滑油磨粒监测试验 ,确定了磨损颗粒各形态参数的聚类权值和灰类白化权函数。应用此方法对一组测试磨粒进行了模拟识别 ,识别正确率在 90 %以上 ,并且识别速度很快 ,大大优于传统的磨粒识别方法。更多还原

【Abstract】 The theory of grey systems is a new technique to perform prediction, relational ana-lysis and decision in many areas. In this paper, the theory of fixed weight grey clustering (FWGC) is used to classify six types of wear particles depended on their morphology numerical parameter analysis. The program of auto-identify of wear particles is carried out by means of FWGC technique after giving out a set of morphology descriptor that describes the micro-feature of wear particles. In the classification process, weight of grey cluster and whitenization weight function for one type of aero engine is studied. A test of classifying the wear particles by this method has been done, and the identifying accuracy is more than 90%, and the identifying speed is very fast. It is much better than the traditional ones.更多还原