摘要
研究一种光纤布拉格光栅传感的轴承异常状态识别方法,用以解决轴承状态精准识别难度过大的问题。使用光纤布拉格光栅传感器采集轴承运行时产生的振动信号,将这些信号频谱使用自适应补偿高斯噪声均匀处理,消除由于经验模态分解算法导致的混叠现象。使用隐半马尔科夫模型分类器对轴承转动初始阶段的振动信号实行训练,获得分类器参数;均匀处理后振动信号作为观测值输入到隐半马尔科夫模型转换动态贝叶斯网络分类器中,得到隐半马尔科夫模型参数,根据参数变化情况,判断轴承状态,实现轴承异常状态识别。结果表明,该方法在识别轴承异常状态时具有时间快、识别率高以及误识别率较低等优势,具有良好的识别效果。
A method for bearing abnormal state identification with fiber Bragg grating sensor was studied to solve the problem that bearing state identification is too difficult. The vibration signals generated during bearing operation are collected by fiber Bragg grating( FBG) sensors,and the spectrum of these signals is uniformly processed by adaptive compensation Gaussian noise to eliminate the aliasing caused by empirical mode decomposition algorithm. The hidden semi-Markov model classifier is used to train the vibration signal in the initial stage of bearing rotation and the classifier parameters are obtained. After the processing,the vibration signal is input as the observed value into the hidden semi-Markov model transformation dynamic Bayesian network classifier,and the parameters of the hidden semi-Markov model are obtained. According to the changes of parameters,the bearing state is judged to realize the identification of bearing abnormal state. The results show that the method has the advantages of fast time,high recognition rate,low error recognition rate,and good recognition effect.
作者
崔雪英
王福顺
崔晶
CUI Xueying;WANG Fushun;CUI Jing(School of Electrical Engineering,Tongling University,Tongling Anhui 244000,China;College of Information Science and Technology,Hebei Agricultural University,Baoding 071000,China;Great Wall Motor Company Limited,Baoding 071000,China)
出处
《激光杂志》
CAS
北大核心
2021年第9期57-61,共5页
Laser Journal
基金
河北省高等学校科学研究项目(No.QN2020421)。
关键词
光纤
布拉格
光栅传感
轴承
异常状态
识别
optical fiber
prague
grating sensing
bearing
abnormal state
identify
作者简介
崔雪英(1976-),女,硕士,讲师,主要研究方向:智能检测与控制;通讯作者:王福顺(1981-),男,博士,副教授,主要研究方向:数据分析与应用。
