Performance assessment and fault classification for hydraulic pump based on LMD and LR
Real-time health monitoring and fault diagnosis system of hydraulic pump is very crucial as the pump being the power source of the entire hydraulic system.Prognostic method based on logistic regression for health assessment and fault classification is proposed.The real-time state of the system is obtained by processing the data of vibration signals collected from the pumps, and maintenance can be performed as long as the failure or malfunction prognosis indicates instead of periodic maintenance inspections.The vibration signal is decomposed into several product functions by local mean decomposition (LMD), and the product functions that contain fault information form a feature vector by abstracting energy values and corresponding time-domain statistical magnitudes.Principle component analysis (PCA) is used for feature reduction.Logistic regression (LR) models are trained by the reduced features to obtain machine health condition and classify possible failure models.The maximum likelihood method is applied to determine the parameters of LR models.The methodology has been applied to process the vibration signals of a real hydraulic pump to verify the effectiveness and feasibility.
hydraulic pump health assessment fault classification local mean decomposition principle component analysis logistic regression model
Yu Ding Jian Ma Ye Tian
Science and Technology on Reliability and Environmental Engineering Laboratory, Beijing, China School of Reliability and Systems Engineering, Beihang University, Beijing, China
国际会议
the International Conference Vibroengineering-2014
贵阳
英文
194-199
2014-11-07（万方平台首次上网日期，不代表论文的发表时间）