文献类型：会议论文

英文题名：Marine Propulsion Shaft Bearing Fault Feature Extraction and Diagnosis Based on Strong Tracking State Principal Component

作者：Liao, Zhiqiang[1,2]; Song, Xuewei[2]; Jia, Baozhu[1]; Zuo, Dunwen[3]; Sheng, Yi[3]; Chen, Peng[2]

机构：[1] Guangdong Ocean University, Maritime College, Zhanjiang, China; [2] Mie University, Graduate School Of Bio-resources, Tsu, Japan; [3] Manufacturing Research Institute, Naning Xinhe Precision Intelligent, Nanjing, China

会议论文集：2021 Global Reliability and Prognostics and Health Management, PHM-Nanjing 2021

会议日期：October 15, 2021 - October 17, 2021

会议地点：Nanjing, China

语种：英文

外文关键词：Extraction - Feature extraction - Ship propulsion - State estimation - Roller bearings - Fault detection - Signal processing

外文摘要：The vibration signal with non-stationary, strong noise, and weak fault feature is inevitably acquired in practical marine propulsion shaft bearing fault diagnosis due to harsh environment. These obstacles lead to diagnostic accuracy degrade and even failure of diagnosis. In light of these problems, a marine propulsion shaft bearing fault feature extraction and diagnosis method based on strong tracking state principal component is presented. Specifically, strong tracking state principal component is employed to build the state model with marine propulsion shaft bearing signal and update the state estimation matrix in each step. The first principal component signal which extracted from state estimation matrix can represent fault feature, and then the extracted first principal component signal is analyzed by envelope demodulation. The dominant frequency in the envelope spectrum is compared with the rolling bearing fault characteristic frequency to fault diagnosis. This presented method is evaluated by simulation signal and practical signal. Moreover, different signal processes methods are selected for comparison, and the comprehensive results validate the effectiveness and superiority of the presented method. ? 2021 IEEE.