摘要
针对滚动轴承信号的非平稳特性、小波变换频率混叠及信息丢失等问题,提出一种结合频率分离与功率谱的滚动轴承故障诊断方法。首先,双树复小波变换(DTCWT)对振动信号进行分解,依据幅频特性对子频段个数进行优化,实现信号频率的精准分离;其次,利用自回归(AR)功率谱得到不同子频段功率,将总功率作为特征输入遗传算法优化的支持向量机(GA-SVM)进行故障诊断。通过实验,复合轴承故障的总体识别率达到96%,其中3种外圈故障识别率达到100%,其结果表明所提方法能够准确提取故障特征并准确识别复合轴承故障。
Aiming at the problems of non-stationary characteristics of bearing signals,frequency aliasing of wavelet transform,and loss of information,a fault diagnosis method for mechanical rolling bearings combining frequency separation and power spectrum is proposed in this paper.First,the dual-tree complex wavelet transform(DTCWT)is utilized to decompose the bearing vibration signal,after that,the number of sub-bands is optimized according to the amplitude-frequency contrast algorithm for the purpose of achieving the separation of signal frequencies accurately.Then,auto-regression(AR)power spectrum is used to get the power of different sub-bands,and the power summation is gained as a feature to the genetic algorithm optimized support vector machine(GA-SVM)for fault diagnosis.Through experiments,the overall fault recognition rate of composite bearing is up to 96%,among which the fault recognition rate of three kinds of outer rings is up to 100%,it shows that the proposed method can accurately extract fault features and accurately identify the fault of composite bearings.
作者
宋玉琴
周琪玮
赵攀
SONG Yu-qin;ZHOU Qi-wei;ZHAO Pan(School of Electronics and Information,Xi′an Polytechnic University,Xi′an 710600,China)
出处
《组合机床与自动化加工技术》
北大核心
2021年第3期31-35,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
中国纺织工业联合会科技指导性项目(2018095)
西安市科技局计划项目(201805030YD8CG14(17))。
关键词
双树复小波变换
幅频特性
自回归功率谱
特征提取
故障诊断
double tree complex wavelet transform
frequency domain parameters
autoregressive power spectrum
feature extraction
fault diagnosis
