Life Prediction and Reliability Analysis of MotorBearing Based on Vibration Signal
As an important power equipment in the power system, the electrialc motor is an important part of the transformation of electric energy and mechanical energy.Reasonable prediction of the reliability level of the running motor, timely and effectively avoid the operational risks, can effectively ensure the safety and stable operation of the power system.This paper starts with the motor vibration signal and establishes the remaining life prediction based on SVM(Support Vector Machine).Firstly, the vibration signal is denoised by wavelet transform, and the characteristic parameters such as mean square value is extracted after signal data reconstruction.Further, combined with the characteristic value, the working state is divided, and the support vector machine regression prediction model is used to obtain the remaining life of the bearing.In the case study part, data denoising and feature extraction are performed on the bearing vibration data measured by the PRONOSTIA test bench.The trend of the selected characteristic values is used for evaluating the motor operating state.The characteristic value, running state of the equipment,and remaining life which are extracted from the training set data are input to the SVM, and then the SVM is used to predict the remaining life for the test set.Finally, failure rate of the equipment can be derived.
Vibration signal wavelet analysis support vector machine life prediction reliability analysis
Helin Xu Lin Cheng Yuxiang Wan Ning Qi
State Key Laboratory of Control and Simulation of Power Systems and Generation Equipment, Tsinghua University, Beijing 100084, China
国际会议
江苏苏州
英文
123-128
2019-11-04（万方平台首次上网日期，不代表论文的发表时间）