摘要
以转炉耳轴轴承为研究对象,将声发射技术应用于转炉耳轴轴承的故障诊断中,提出了应用主成分分析(PCA)和最小二乘支持向量机(LS-SVM)相结合的故障诊断方法。首先,对声发射信号的特征量进行主成分分析,得到更能反映设备状态的综合特征参数,然后将新的特征参数输入到最小二乘支持向量机中进行状态识别。利用在实际生产过程中采集到的转炉耳轴轴承声发射数据进行方法验证。结果表明,新方法能够有效区分出转炉耳轴轴承的故障模式,识别的总体正确率可达97.8%。
Taking converter trunnion bearing as a research object, Acoustic Emission （AE） technology is applied to the fault diagnosis for the converter trunnion bearings by using a new method which employs the combination of Principle Component Analysis （PCA） and Least Squares Support Vector Machines（ LS -SVM）. Firstly, the AE features are cal- culated, and then features are extracted by PCA to get the comprehensive feature parameters. The results are put into LS - SVM to achieve the pattern recognition. AE signals are acquired from the converter trunnion bearings in production condition. The results show that the proposed method is effective in distinguishing fault modes of the converter trunnion bearings, and the accuracy of recognition is 97.8%.
出处
《轴承》
北大核心
2013年第1期46-50,共5页
Bearing
基金
国家自然科学基金资助项目(50905013
51004013)
高等学校博士学科点专项科研基金项目(20090006120007)
关键词
转炉耳轴轴承
故障诊断
最小二乘支持向量机
主成分分析
声发射
converter trunnion bearing
fault diagnosis
least squares support vector machine
principle component a-nalysis
acoustic emission
