基于颜色特征提取的磨粒材质识别

Wear Debris Material Recognition Based on Color Feature Extraction

在机械设备磨损过程中,不同材质摩擦副磨损产生的磨粒颜色各不相同,通过研究磨粒颜色可以判断发生磨损的材质及磨损产生机制,进一步判断出磨损或故障所发生的部位。采用典型材质磨粒颜色标准库,构建基于磨粒颜色模糊辨识的材料判断方法,利用K-Means聚类算法和基于欧氏距离的最小距离分类法实现磨粒颜色提取及典型材质的自动识别,为机器设备磨损故障监测中的机制分析及磨损部位判定提供了新方法。

In the wear process of mechanical equipment,the color of the wear particles produced by friction pairs of different materials is different.By studying the color of the wear debris,the material of wear and the mechanism of wear can be deduced,further the wear or failure parts can be determined.A material judgment method based on fuzzy identification of abrasive color was constructed by using the standard database of abrasive color of typical materials.The extraction of abrasive color and automatic identification of typical materials were realized by using K-Means clustering algorithm and the minimum distance classification method based on European distance,which provides a new method for mechanism analysis and determination of wear parts in wear fault monitoring of machine equipment.