Research On Rayleigh Wave Signal De-noising And Dispersive Curve Extraction In Tunnel Detecting

Advanced geological prediction plays a vital role in the tunnel construction. In most tunnels construction projects, it is the indispensable technology. As Rayleigh wave exploration has many advantages, such as low equipment cost, wide application range, less affected by the site, simple detection equipment and fast detection speed and so on, it is introduced in advanced geological prediction for highway tunnels. Based on this, the dissertation studies Rayleigh wave dispersion curve forward modeling, Rayleigh wave signal de-noising and purification and the dispersion curve extraction.According to the typical tunnel geological anomalies i.e. fault fracture zone, it designs three categories, a total of seven kinds of model. Based on derivation and solution of the Rayleigh wave dispersive equation, it carries out forward modeling research on the tunnel geological anomalous dispersion curve. Research shows that the anomaly of Frontage of Heading will lead to the ’zigzag’ inflection point of Rayleigh wave dispersion. The take-off point of the’zigzag’ reflects the position of the geological anomaly.’Zigzag’ deformation span has great concern with the position, size scale and nature of geological anomaly. The shallower the geological anomaly buries and the bigger the size of the geological anomaly is and the more obvious low-speed characteristics is, the greater the deformation of the ’zigzag’ spans. It proposes the applicable measuring point layout of Rayleigh wave exploration and data acquisition method after analyzes the application principles of Rayleigh wave in advanced geological prediction and combines the engineering practices.Since the effect of the superposition of signal de-noising is not good enough in the original instrument, it uses the wavelet analysis to make research on Rayleigh wave signal de-noising. It respectively leads in two methods that are wavelet threshold and wavelet modulus maximum for the research of Rayleigh wave signal de-nosing. Through a lot of simulation experiments and engineering practices, the research shows that wavelet threshold method can de-noise high frequency random noise signal, but has little effect on removing the interference wave. Wavelet modulus maximum method not only can remove the random noise signal, but only can filter the interference wave according to the energy features of Rayleigh wave.Considering that Rayleigh wave and other interference wave have differences on energy features, spreading speed, duration and signal correlation, it proposes a new Rayleigh wave purification method based on empirical mode decomposition (EMD). Based on the basic principle analysis and derivation of empirical mode decomposition, it establishes the algorithm of Rayleigh wave signal purification on EMD and makes a program. Research shows that it can obtain different physical intrinsic mode function (IMF) after decomposing the records by EMD, which can effectively distinguish Rayleigh wave from other interfering signal and purify the Rayleigh wave ingeniously. The research is proved by engineering example testing, and it also points out this method is adaptable, stable and reliable.Since the tunnel face horizontal width is limited, it cannot use multichannel Rayleigh wave exploration technology. It applies spectral analysis of surface waves (SASW) to purify dispersion curve by two channels. According to the low resolution of its existing deep exploration, it puts forward improved algorithm by increasing the sampling points and develops the program. The improved algorithm can enhance resolution in deep part and the recognition of geological anomaly.According to the wavelet transform along with the time domain-frequency domain resolution characteristics, it breaks the bottlenecks of Fourier transform methods and introduces a new method to extract dispersion curve on wavelet transform. Through theoretical calculations, it establishes the algorithm of extraction dispersion curves and develops a program. After experimenting and analyzing the engineering examples, it shows that wavelet transformation method is a very effective method to calculate Rayleigh wave dispersion curve and it has the advantages on good purification performance, reliable, partly high resolution and high precision in deep detection.