基于EEMD的地铁隧道爆破振动信号分析与应用研究

针对EMD（empirical mode decomposition，经验模态分解）方法在处理爆破振动信号时存在模态混叠的问题，引入EEMD（ensemble empirical mode decomposition，集合经验模态分解）方法，对爆破振动信号进行去噪处理，并分析城市隧道掘进爆破振动信号的特征信息，从而研究城市隧道掘进爆破振动效应。以乌鲁木齐轨道交通1号线爆破开挖工程为依托，利用现场爆破振动监测数据，采用EEMD方法与EMD方法对原始爆破振动信号进行去噪处理，运用信噪比（SNR）法和均方根误差（RMSE）法对所得结果进行量化评估并进行比较，从而获得较准确的信号去噪方法，然后运用希尔伯特变换对去噪后的爆破振动信号进行时间-频率和能量分布特征分析。研究表明： 1）EEMD法去噪结果信噪比和均方根误差分别为30.51和0.005 5，EMD法去噪结果信噪比和均方根误差分别为20.88和0.010，EEMD去噪法要优于EMD去噪法； 2）在文章所述起爆方式下，隧道爆破振动的能量一般集中在频率段5～80 Hz，且主要分布在50 Hz以下的低频段，应采取合理的降震增频措施来控制爆破振动对建（构）筑物的影响； 3）爆破振动瞬时能量峰值与振速峰值所在的时刻一致，说明瞬时能量谱能较好地体现爆破过程中振动效应随时间的变化规律； 4）基于信号去噪及分析的研究结论，提出地铁隧道爆破振动控制技术。

Analysis and Application of Metro Tunnel  Blasting Vibration Signal Based on EEMD

When using empirical mode decomposition(EMD) to process blasting vibration signal, the problem of mode aliasing occurs. Hence, the ensemble empirical mode decomposition(EEMD) method is introduced to denoise the blasting vibration signal; and the characteristic information of blasting vibration signals in urban tunnel excavation is analyzed to study the vibration effect of urban tunnel blasting excavation. By taking the blasting excavation project of Urumqi Rail Transit Line 1 for example, the original blasting vibration signals are denoised by EEMD and EMD methods; the single noise ratio(SNR) method and the root mean square error(RMSE) method are used to quantitatively evaluate and compare the results to gain a more accurate signal denoising method; and the time frequency and energy distribution characteristics of the denoised blasting vibration signal are analyzed by Hilbert transform. The study results show that: (1) The SNR value and RMSE value of denosing result by EEMD method are 30.51 and 0.005 5, respectively, and those of denosing result by EMD are 20.88 and 0.010, respectively, which show that the denosing effect of EEMD method is better than that of EMD method. (2) In current blasting operation, the energy of blasting vibration is mainly distributed in frequency section of 5-80 Hz, mainly in frequency section below 50 Hz; hence, proper measures should be taken to control the impact of blasting vibration on buildings or structures. (3) The instantaneous energy peak value of blasting vibration coincides with the peak value of vibration velocity, which indicates that the instantaneous energy spectrum can better reflect the variation law of vibration effect with time in blasting process. (4) Based on the research conclusion of signal denoising and analysis, the blasting vibration control technology of metro tunnel is put forward.