摘要
在磁瓦内部缺陷声振检测中,具备可预设尺度和自适应能力的变分模态分解(Variational Mode Decomposition,VMD)在处理相关声振信号时具有明显优势。然而VMD的参数选择范围广且对分解效果影响大,为解决参数的统一预设问题,提出一种基于天牛须搜索(Beetle Antennae Search,BAS)的VMD参数优化方法,并应用于磁瓦内部缺陷检测。该方法首先根据声振信号特点建立基于模态能量和模态中心频率变化规律的目标函数;再利用BAS在参数空间中搜寻目标函数最大值,所需的VMD统一预设参数即为该最大值所对应的参数组合;最后,在该优化参数下对逐个信号进行VMD处理并提取模态中心频率特征,进而由支持向量机完成特征识别。试验结果表明,所提出的方法能有效优化VMD参数并实现磁瓦内部缺陷的快速精准检测。
Variational mode decomposition(VMD)with preset scale and adaptive ability has significant advantages in processing acoustic-vibration signals for detecting internal defects of magnetic tile,but VMD parameters’selection ranges are wide and have large influences on its decomposition effect.Here,in order to solve the problem of uniform preset of VMD parameters,a novel VMD parametric optimization method based on the beetle antennae search(BAS)algorithm was proposed.Firstly,an objective function based on mode energy distribution and mode center frequency variation was established according to acoustic-vibration signal’s features.Secondly,BAS was used to search the maximum value of the objective function in VMD parametric space.The parameters corresponding to the maximum value were taken as the optimal VMD parameters.Finally,each acoustic-vibration signal was processed using VMD with the optimal parameters to extract mode center frequencies’features,and further complete feature recognition with support vector machine.Test results showed that the proposed method can effectively optimize VMD parameters,and realize a rapid and accurate detection of internal defects of magnetic tile.
作者
黄沁元
谢罗峰
殷国富
冉茂霞
刘鑫
HUANG Qinyuan;XIE Luofeng;YIN Guofu;RAN Maoxia;LIU Xin(School of Automation and Information Engineering,Sichuan University of Science and Engineering,Zigong 643000,China;School of Manufacturing Science and Engineering,Sichuan University,Chengdu 610065,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2020年第17期124-133,共10页
Journal of Vibration and Shock
基金
国家自然科学基金项目(61701330)
四川省教育厅科研项目(18ZB0428)
人工智能四川省重点实验室开放基金项目(2016RZJ01)
四川理工学院人才引进项目(2016RCL29)。
关键词
变分模态分解(VMD)
天牛须搜索(BAS)
磁瓦
内部缺陷
声振信号
variational mode decomposition(VMD)
beetle antennae search(BAS)
magnetic tile
internal defect
acoustic-vibration signal
