Mechanical Fault Prediction Based on Principal Component Analysis

摘要：

In order to find an effective way to predict running of mechanical equipment, a new prediction method of mechanical fault based on principal component analysis (PCA) is proposed in this study. The disadvantages of traditional prediction methods and the presence and development in mechanical fault of PCAbased predication method were briefly introduced. Theoretical basis, analysis processes and parameters selection of PCA were also investigated. It is proved that PCA-based prediction neglects the structure and principle of system in detail and only depends on the data from the sensor. Simulation experiments show that the PCA-based prediction method is a promising and feasible way to forecast mechanical failure, because the algorithm simple and easy to implantation, and it can reduce noise and simplify data processing.

关键词： principal component analysis data-driven fault prediction

作者: Luhui Lin Jie Ma Xiulan Ye Xiaoli Xu

作者单位: College of Automation,Measurement & Control Technology (Ministry of Education)Beijing Information Sc Beijing Navy 701 Factory No.3,Jiuxianqiao Road,Chaoyang DistrictBeijing,100016,China Key Laboratory of Modern Measurement & Control Technology (Ministry of Education) Beijing Informatio

会议类型: 国际会议

会议名称: 2010 IEEE信息与自动化国际会议(ICIA 2010)

会议地点: 哈尔滨

会议语种:英文

页码: 1-5

在线出版日期: 2010-06-20（万方平台首次上网日期，不代表论文的发表时间）