爆破振动信号3种经验模态分解差异性研究

Study on difference of three empirical mode decompositions of blasting vibration signal

为了准确评估冻结立井爆破对井壁产生的影响,采用井壁预埋法对大药量爆破下井壁的振动响应进行了监测。利用EMD、EEMD和CEEMD典型经验模态算法对井壁信号进行了分析,并结合时频谱对分解和重构效果进行了综合评价。分析结果表明:受测试环境影响,爆破信号中普遍含有噪声等干扰成分。EMD分解存在明显模态混叠和端点效应,EEMD分解虽对模态混叠现象有所改善,但去噪效果仍不理想,CEEMD分解对模态混叠和噪声消除方面均具有很好的处理效果。CEEMD重构信号时频谱能够深刻揭示爆破能量在时频域上的分布且对干扰成分不敏感,适合用于批量信号的预处理过程。分析结果对于爆破能量识别和振动损伤控制具有积极的现实意义。

In order to accurately evaluate the influence of freezing shaft blasting on shaft lining,the vibration response of shaft lining under large charge blasting is monitored by using the shaft lining embedding method.EMD,EEMD and CEEMD are used to decompose and reconstruct the blasting signal,combined with the time-frequency spectrum,and the effects of three methods are evaluated comprehensively.The analysis results show that:affected by the test environment,blasting signal generally contains noise and other interferences.There are obvious mode aliasing and endpoint effects in the EMD method.The EEMD method can improve the mode aliasing,but the denoising effect is still not ideal.CEEMD method has a good effect on both mode aliasing and noise elimination.CEEMD reconstructed signal time frequency spectrum can reveal the distribution of blasting energy in time frequency domain,and is not sensitive to interference elements,so it is suitable for batch pre-processing of signals.The analysis results have positive practical significance for blasting energy identification and vibration damage control.