基于CEEMD与TQWT组合方法的爆破振动信号精细化特征提取

摘要

针对传统小波在爆破振动信号特征提取和分析方面的局限性，提出了基于 CEEMD 和 TQWT 组合的信号精细化特征提取方法。预先设定可调品质因子小波 TQWT 高、低品质因子参数对 CEEMD 分解优势分量重组信号进行分解，并引入相对权重因子θ，优化了分解过程，实现了爆破振动信号特征的精细化提取。分析结果表明：组合方法对爆破振动信号的分析不依赖于先验小波基的选择，分解过程实现了信号的二次滤波。通过连续小波多尺度三维谱和时频小波脊线对比，说明组合算法分解得到的最佳分析信号可真实反映振动信号的细节信息，时频分辨率更高。该组合方法抑制了杂波分量对信号特征的干扰，可精确地提取复杂环境下的爆破振动信号特征信息。%Aiming at the limitations of traditional wavelet transformation in feature extraction and analysis of blasting vibration signals,a combination method of Tunable Q-factor wavelet Transform (TQWT)and Complementary Ensemble Empirical Mode Decomposition (CEEMD)signal refine feature extraction was proposed.Presetting tunable wavelet TQWT high and low quality factors’parameters,the reconstructed signals obtained with dominant components of CEEMD were decomposed.The relative weighted factor θwas introduced to optimize the decomposition process.The fine feature extraction of blasting vibration signals was realized.Analysis results showed that the combination method is not dependent on choices of priori wavelet basis to analyze blasting vibration signals,two filterings of signals are realized in the decomposition process;through comparing continuous wavelet multi-scale 3D spectrum and time frequency wavelet ridge lines,the optimal analysis signals obtained with the decomposition of the combined algorithm can truly reflect the details of vibration signals,the time-frequency resolution rate is higher;the combined method suppresses the interference of clutter wave components to signals’characteristics,it can extract accurately the characteristics information of blasting vibration signals under complex environment.