摘要
针对机械故障中产生的磨损情况,通过磨粒识别可以有效的提高设备的故障诊断和监测水平,减少机械故障事故发生的概率。文中对难分析的氧化物磨粒、严重滑动磨粒、疲劳磨粒提出了针对性的识别方法。提出利用主成分分析与欧氏距离相结合的方法识别红色氧化物磨粒和黑色氧化物磨粒;灰色关联分析和主成分分析相结合的方法识别严重滑动磨粒和疲劳磨粒,最后作者通过实例,验证了上述方法的准确性和可行性,提高了磨粒识别的速度和效率。
According to abrasive wear of the mechanical failure,abrasive recognition technology can be used to effectively improve the equipment fault diagnosis and monitoring of standards and reduce the occurrence of mechanical failure.Identification of specific analysis methods has been proposed oxide abrasive,abrasive severe sliding,fatigue,abrasive.Principal component analysis combined with the Euclidean distance identification oxide abrasive.Grey relational analysis and principal component analysis combined analysis identified fatigue and severe sliding abrasive.Finally,verified the accuracy and feasibility of the method by example,abrasive identification speed and efficiency is improved.
出处
《计算机技术与发展》
2012年第4期16-20,共5页
Computer Technology and Development
基金
南京航空航天大学基本科研业务费专项科研项目(NS2010136)
关键词
磨粒识别
主成分分析
灰色关联度
欧氏距离
wear particles identification
principal component analysis
grey relation degree
Euclidean distance
