摘要
针对复杂机械振动信号压缩感知过程中存在的稀疏字典构造困难问题,提出了基于QPSO分类的自适应稀疏字典构造方法。该方法根据信号的分割尺度,将信号进行分块,并利用每一信号块的能量大小,构造能量序列,利用QPSO对能量序列进行优化分类,保证不同类别间能量序列的方差最大,从而实现对信号块的分类,采用K-SVD对不同类信号块分别进行稀疏字典的自适应学习训练,产生与信号相适应的稀疏字典,用于机械振动信号的压缩感知重构过程。通过滚动轴承实测信号在不同状态下的压缩感知实验表明:所提方法能够有效提高信号重构的峰值信号比,改善机械振动信号的重构效果。
Aiming at the difficulty of dictionary construction for machinery vibration signals in the process of compressed sensing, an adaptive dictionary construction algorithm based on signal classification was put forward.Machinery vibration signals were divided into blocks according to their cutting size,and an energy sequence was produced in accordance with the energy of each signal block. Using the QPSO algorithm,the energy sequence was classified where the variance of the sequence between different classes was ensured to be the biggest. Then the classification of the signal blocks was realized by virtue of the energy sequence. Finally,the dictionary of different class of signal blocks was constructed by using the mcthod of K-SVD,and the reconstruction of machinery vibration signals was achieved. The experiments on rolling bearing in different status show that the method put forward can increase the peak signal to noise ratio of reconstructed signals and improve the effects of reconstruction for machinery vibration signals.
作者
王强
张培林
王怀光
吴定海
张云强
WANG Qiang;ZHANG Peilin;WANG Huaiguang;WU Dinghai;ZHANG Yunqiang(Department of Vehicle and Electrical Engineering,Ordnance Engineering College,shijiazhuang 050003,Chin)
出处
《振动与冲击》
EI
CSCD
北大核心
2018年第14期86-93,共8页
Journal of Vibration and Shock
基金
国家自然科学基金(E51305454)
关键词
机械振动信号
字典构造
自适应压缩感知
优化分类
machinery vibration signals
dictionary construction
adaptive blocked compressed sensing
optimal classification
