隧道瑞雷波探测信号去噪与频散曲线提取研究

发布时间：2018-07-16 08:58

【摘要】：隧道超前地质预报对于保障隧道施工安全起到了关键性的作用,绝大部分隧道工程项目都将超前地质预报列为必要的技术环节。由于瑞雷波勘探具有设备成本低、受场地限制小、应用领域广、检测快速方便等优点,因此把瑞雷波勘探引入公路隧道施工超前地质预报,开展了隧道瑞雷波频散曲线正演、信号去噪与提纯、频散曲线提取等方面的研究。根据隧道典型地质异常即断层破碎带设计了三个类别共七种模型,通过对瑞雷波频散方程的推导和求解,开展了隧道地质异常频散曲线正演研究。研究表明隧道掌子面前方的异常带会使瑞雷波频散曲线产生“之”字形拐点,“之”字形的起跳位置基本反映了地质异常的位置,且“之”字形变形跨度与地质异常的位置、尺寸规模和性质都有很大的关系,地质异常埋深越浅、地质异常的尺寸规模越大、低速特性越明显,则之字形变形跨度越大。分析了瑞雷波勘探技术应用于隧道超前地质预报的原理,结合工程实践,提出瑞雷波勘探合适的测点布置及数据采集方式。针对原仪器采用信号叠加去噪的效果不良的情况,基于小波分析原理进行瑞雷波信号去噪研究,分别利用小波阈值及小波模极大值两种方法开展去噪研究,通过仿真实验和工程实例,指出小波阈值方法能够很好地去除高频随机噪声,但对于去除干扰波并没有突出的作用,而小波模极大值方法不但能够有效去除随机噪声,而且根据瑞雷波的能量特征,能够起到滤除干扰波的效果。考虑到瑞雷波和其他干扰波在能量特征、传播速度、持续时间及信号相关性等多方面的存在特征区别,提出一种基于经验模态分解(EMD)的瑞雷波信号提纯新方法。通过对经验模态分解基本原理理论分析和推导,建立瑞雷波信号EMD提纯的实施算法并编制程序。研究表明勘探记录信号通过EMD分解后,得出了具有不同物理意义的本征模态函数(IMF),使瑞雷波信号和其他干扰信号能够有效区分开来,从而巧妙地将瑞雷波信号提纯出来,并通过了工程实例验证。指出该方法具有自适应性、稳定性和可靠性。由于隧道掌子面横向净宽有限,无法采用多道瑞雷波技术。应用基于两道技术的表面波谱法提取频散曲线,根据其存在的深部探测分辨率不高的不足,提出增加采样点的改进算法并编制出程序。改进后的算法能够提高深部分辨率,增强了地质异常的识认性。根据小波变换同时具备时域-频域分辨的特点,突破原有方法均采用傅里叶变换的瓶颈,提出了基于小波分析的频散曲线提取新方法,通过理论推导建立了小波分析提取频散曲线的算法,并编制程序,经实验和工程实例研究分析表明,小波变换法是计算瑞雷波频散曲线的一种非常有效的方法,且具有提纯性能好、可靠性高、局部分辨率高、深部探测精度高的优点。
[Abstract]:The advance geological prediction of tunnel plays a key role in guaranteeing the safety of tunnel construction. Most of the tunnel projects have listed ahead geological forecast as the necessary technical link. Because of the advantages of low cost of equipment, small limit of site, wide application field, rapid and convenient inspection and so on, Leibo exploration leads to the exploration of Leibo. The advance geological forecast of highway tunnel construction has been carried out, and the research of tunnel Rui Leibo frequency dispersion curve forward, signal denoising and purification, frequency dispersion curve extraction, and so on. According to the typical geological anomaly of the tunnel, three categories of seven kinds of models are designed. The tunnel geology is carried out through the derivation and solution of the frequency dispersion equation in Leibo. The study shows that the abnormal band in front of the tunnel face will make the curve of the Leibo frequency dispersion curve "the" character inflection point, and the take-off position of the "it" shape basically reflects the location of the geological anomaly, and the shape of the "it" shape span is closely related to the location of geological anomalies, size scale and nature. The more shallow the abnormality of the mass, the larger the size of the geological anomaly, the more obvious the characteristics of the low speed, the greater the span of the shape deformation. The principle of the application of Leibo exploration technology to the advance geological prediction of the tunnel is analyzed, and the suitable location and data acquisition mode of the exploration in Leibo are put forward in combination with the engineering practice. In the case of bad de-noising, based on the principle of wavelet analysis, we study the de-noising of Leibo signal, and use the two methods of wavelet threshold and wavelet modulus maximum to carry out denoising. Through simulation experiments and engineering examples, it is pointed out that the wavelet threshold method can remove the high frequency random noise well, but it is not sudden to remove the interference wave. The wavelet modulus maxima method not only can effectively remove the random noise, but also can filter the interference wave according to the energy characteristics of Leibo, taking into account the characteristics of the energy characteristics, the propagation speed, the duration and the signal correlation of the other interference waves, which are based on the characteristics of the Leibo and other interference waves. The empirical mode decomposition (EMD) is a new method to purify the Swiss Leibo signal. Through the theoretical analysis and deduction of the basic principle of empirical mode decomposition, the implementation algorithm of EMD purification is established and the program is compiled. The study shows that the exploration record signal is decomposed by EMD, and the intrinsic modal function (IMF) with different physical significance is obtained, so that Rayleigh is made. The wave signal and other interference signals can be distinguished effectively, so that the Leibo signal is refined and verified by an engineering example. It is pointed out that the method has adaptability, stability and reliability. Because the transverse width of the tunnel face is limited, the multi road Leibo technology can not be used. The surface wave based on the two channel technology is applied. The spectral method is used to extract the frequency dispersion curve. According to the lack of high resolution in the deep detection, the improved algorithm and program are put forward. The improved algorithm can improve the resolution of the depth and enhance the recognition of the geological anomaly. With the bottleneck of Fourier transform, a new method of dispersion curve extraction based on wavelet analysis is proposed. The algorithm of extracting dispersion curve by wavelet analysis is established through theoretical deduction and program is compiled. It is shown by experiments and engineering examples that the wavelet transform is a very effective method for calculating the dispersion curve of Leibo. It has the advantages of good purification performance, high reliability, high local resolution and high accuracy of deep detection.
【学位授予单位】：福州大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U452.11

【相似文献】