改进的VMD⁃WT微震信号联合去噪方法

Improved VMD-WT Microseismic Signal Joint Denoising Method

摘要:

外部环境噪声信号的存在，影响着微震监测系统对岩体破裂灾害的预警效果。针对微震信号具有非线性、随机性强、非稳定的特点与传统VMD、WT算法中在去噪处理时存在一定局限性的问题，提出了一种改进的VMD- WT联合去噪方法。首先，使用GSWOA算法对VMD中的分解个数及惩罚因子进行参数寻优，将优化后的参数代入VMD算法中将含噪信号分解为若干个IMF分量；其次，使用MI法对IMF分量进行分类，将有效分量保留并重构信号；最后，使用GSWOA算法对改进阈值函数的WT算法进行参数寻优，实现对含噪信号的二次去噪。对构建的仿真信号进行去噪处理，验证了改进后的联合去噪方法的可行性与优越性；并进一步将此方法应用于实测微震信号的去噪处理中，并以信噪比、均方根误差、平方绝对误差作为去噪效果评价指标，结果表明，与单一的EMD、 WT、VMD去噪算法及EMD-SVD、VMD-SVD联合去噪算法相比，改进的VMD-WT去噪方法能在保留原有信号信息的基础上，更好地去除微震信号中的噪声干扰，为后续利用微震监测系统对岩体破裂灾害进行预警奠定基础。

Abstract:

External environmental noise signals affect the early warning performance of microseismic monitoring systems for rock mass rupture disasters. An improved VMD-WT joint denoising method was proposed to address the nonlinear, highly random, and unstable characteristics of microseismic signals, along with the limitations of traditional VMD and WT algorithms in denoising. First, the GSWOA algorithm was used to optimize the decomposition number and penalty factor in the VMD process. The optimized parameters were substituted into the VMD algorithm to decompose the noisy signal into several IMF components. Next, the MI method classified the IMF components, retained the effective components, and reconstructed the signal. Finally, the GSWOA algorithm optimized the parameters of the improved threshold function in the WT algorithm for secondary denoising of the noisy signal. The feasibility and superiority of the improved joint denoising method were verified by denoising simulated signals. The method was further applied to real microseismic signals. Its denoising performance was evaluated using signal-to-noise ratio (SNR), root mean square error (RMSE), and mean square error (MSE). The results showed that, compared to the individual EMD, WT, and VMD denoising algorithms, and the EMD-SVD and VMD-SVD joint denoising methods, the improved VMD-WT method more effectively removed noise interference from microseismic signals while preserving the original signal information. This method provides a solid foundation for future early warning of rock mass rupture disasters using microseismic monitoring systems.