煤岩声发射信号识别研究

Research of identification of acoustic emission signal of coal and rock

针对煤岩破裂的声发射信号难以在复杂的噪声环境中识别的问题,提出了一种基于小波包分析和小波特征能谱系数分析的煤岩声发射信号识别方法。选取Symlets小波作为煤岩声发射信号分析的小波基函数,采用混合阈值算法对该信号进行去噪处理,提取出有用声发射信号,并采用Matlab软件分别对有用声发射信号和噪声信号的小波包分解进行仿真,得到两者的小波特征能谱系数和小波包特征向量。仿真结果表明,有用声发射信号特征向量的各级能量变化程度较大,噪声信号特征向量的能量变化较为稳定,从而可实现煤岩声发射信号的识别。

In view of problem that it is difficult to identify acoustic emission signal of coal and rock burst under complicated noise environment, the paper proposed an identification method of acoustic emission signal of coal and rock based on wavelet packet and wavelet feature energy spectrum coefficient analysis. Useful acoustic emission signals are extracted by taking Symlets wavelet as wavelet basis function of acoustic emission signal of coal and rock and making denoising process with hybrid threshold algorithm. Then wavelet feature energy spectrum coefficient and wavelet packet eigenvector are obtained by using Matlab software to separately simulate wavelet packet decomposition for the useful acoustic emission signals and noise signals. Simulation results show that changing degree of each energy of eigenvector of the useful acoustic emission signals is bigger, while changing of energy of eigenvector of the noise signals is relatively stable, which can be used to identify acoustic emission signal of coal and rock.