基于磨损机理的磨粒图像识别仿真  

Recognition Simulation of Wear Particle Image Based on Wear Mechanism

作　　者：杨文君 孙耀宁[1] 梁国强 王雅[1] YANG Wen-jun;SUN Yao-ning;LIANG Guo-qiang;WANG Ya(Department of Mechanical Engineering,Xinjiang University,Urumqi Xinjiang 830047,China)

机构地区：[1]新疆大学机械工程学院,新疆乌鲁木齐830047

出　　处：《计算机仿真》2020年第2期459-462,共4页Computer Simulation

基　　金：国家自然科学基金资助项目(51465055)。

摘　　要：机械设备磨损过程中产生的磨粒,可以利用智能识别技术进行识别。通过对切削磨粒、球状磨粒、疲劳磨粒以及严重滑动磨粒的磨损机理的研究,提出了能够识别各类磨粒的显著特征,将特征参数进行量化表征,并以特征参数为输入向量,建立支持向量机分类器模型,运用层次法对分类器进行训练,优化分类器的参数,最后利用分类器模型对磨粒图像进行识别以验证识别方法的可行性。实验结果表明,支持向量机分类器识别磨粒类型准确率较高,可以用于磨粒图像的识别。Abrasive particles generated during the wear of mechanical equipment can be identified using intelligent recognition technology.Through the study of the wear mechanism of cutting abrasive grains,spherical abrasive grains,fatigue abrasive grains and severe sliding abrasive grains,it is proposed to recognize the distinctive features of various abrasive grains,quantify the characteristic parameters,and take the characteristic parameters as input vectors.The support vector machine classifier model is established.The classifier is trained by the hierarchical method to optimize the parameters of the classifier.Finally,the classifier model is used to recognize the abrasive image to verify the feasibility of the recognition method.Simulation results show that the support vector machine classifier can recognize the abrasive grain type with high accuracy and can be used for the recognition of abrasive image.

关 键 词：数字化特征 磨粒识别 支持向量机 

分 类 号：TH117.1[机械工程—机械设计及理论]