基于小波包分析的含水岩石破裂前兆研究

发布时间：2018-01-14 21:18

本文关键词：基于小波包分析的含水岩石破裂前兆研究　出处：《东北大学》2014年硕士论文　论文类型：学位论文

【摘要】：材料在应力作用下的变形与裂纹扩展是结构失效的重要机制,当材料局部区域产生应力集中现象时,能量会被瞬间快速释放。因此,从能量的角度来研究含水岩石失稳破坏的性质是切实可行的。然而,当这种能量以声发射信号的方式被接收到的过程中,由于其非平稳性,使得利用传统的信号分析技术难以对获得的声发射信号进行全面、系统、有效的解读。因此,通过何种方法能更为行之有效的解读与含水岩石破坏过程如影相随的能量变化便成了各国专家与学者争相研究的热点问题。论文首先在实验室内通过对不同饱和度的岩石进行单轴压缩试验,获取其破坏过程中的声发射数据,然后,研究了随着饱和度的变化,砂岩声发射事件数的变化规律。其次,利用小波包变换对声发射信号进行处理,在比较了利用小波分析和小波包分析两种降噪方法的优劣性的同时较为有效的对信号进行了降噪。通过对小波包能量频带的分析,研究了含水岩石声发射信号的能量频带分布特点,比较了在破裂前后能量频带的变化规律,寻找到了含水岩石破坏过程中的主频带。再次,根据能量原理,依据时间序列,利用信号奇异性,通过小波包分解与重构,计算Lipschitza值,将最小α值对应的时间序列中的时间点作为含水岩石破裂的时间,较为有效地提取到了含水岩石破裂的时间信息。最后,根据一个矩阵可由它的所有特征向量完全表示,而每一个向量所对应的特征值,就代表了矩阵在这一向量上的贡献率的理论,基于当信号发生变化时,信号的各频率成分间会产生不同的抑制或增强作用,与非破裂的信号相比,破裂前一小段时间内的信号上的某些频带的能量会激增,某些频带上的能量会骤减的思想,将声发射信号进行小波包分解并重构得到小波包结点系数,以小波包结点系数为特征向量,求得小波包结点系数特征值,通过特征值的突变来预测含水岩石的破裂。虽然实验室内的岩样与工程中的岩体的声发射信号的能量间有所差别,但是万变不离其宗,本文对含水岩石的损伤过程和发展趋势及破裂预测具有现实的指导意义。
[Abstract]:The deformation and crack propagation of materials under stress is an important mechanism of structural failure. When the stress concentration occurs in the local region of materials, the energy will be released instantly and quickly. It is feasible to study the instability and failure of water-bearing rock from the angle of energy. However, when this energy is received as acoustic emission signal, it is due to its non-stationarity. It is difficult to use the traditional signal analysis technology to interpret the acoustic emission signal comprehensively, systematically and effectively. What method can be used to understand the energy change with the destruction process of water-bearing rock has become a hot issue for experts and scholars all over the world. The uniaxial compression test was carried out for the rock of degree degree. The acoustic emission data in the process of destruction are obtained, and then the change rule of the number of sandstone acoustic emission events with the change of saturation is studied. Secondly, the wavelet packet transform is used to process the acoustic emission signal. The advantages and disadvantages of wavelet analysis and wavelet packet analysis are compared. The characteristics of energy band distribution of acoustic emission signal of water-bearing rock are studied. The variation law of energy frequency band before and after rupture is compared and the main frequency band in the process of failure of water-bearing rock is found. Thirdly, according to the principle of energy. According to the time series and signal singularity, the Lipschitza value is calculated by wavelet packet decomposition and reconstruction, and the time points in the time series corresponding to the minimum 伪 value are taken as the time of water-bearing rock fracture. Finally, according to a matrix can be completely represented by all its eigenvector, and each vector corresponding to the eigenvalue. This theory represents the contribution rate of matrix on this vector, based on the fact that when the signal changes, the frequency components of the signal will have different suppression or enhancement effects, compared with the non-broken signal. The energy of some frequency bands will surge in a short period of time before rupture, and the energy in some frequency bands will be suddenly reduced. The acoustic emission signal will be decomposed into wavelet packets and reconstructed to obtain the wavelet packet node coefficients. The eigenvalue of the wavelet packet node coefficient is obtained by using the wavelet packet node coefficient as the eigenvector. Although the energy of acoustic emission signals of rock samples in laboratory is different from that of rock mass in engineering, the variation of rock samples is not different from that of rock samples in engineering. This paper has practical guiding significance for damage process, development trend and fracture prediction of water-bearing rock.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU45

【共引文献】