Application Of Adaptive Time-frequency Analysis In Ground Penetrating Radar Signal

The application of geophysical methods in environmental pollution monitoring is mainly based on the difference of physical and chemical properties between pollutants and their surrounding media and the study and analysis of physical laws,useing special equipment and instruments to measure the distribution of pollutant physical field.Then combined with the relevant data,the distribution characteristics of pollutants in a certain underground depth range are inferred and explained.Electrical prospecting has become a geophysical method widely used in environmental pollution investigation,monitoring and prevention.Ground penetrating radar is one of the most widely used electromagnetic detection methods in environmental and engineering exploration.In traditional ground penetrating radar signal analysis,the signal is generally considered to be stationary and linear,so the traditional ground penetrating radar signal analysis method is Fourier analysis.For non-stationary non-linear signals,because Fourier transform transforms the whole time domain of the signal into frequency domain,all time domain components are used to reflect the change of the signal in frequency domain and the local transient characteristics of the signal can not be reflected.In time-frequency analysis,signal energy is a two-dimensional function of time and frequency.The common disadvantage of integral time-frequency analysis methods is that they are constrained by the lack of adaptability of Fourier transform.The time-frequency analysis method based on empirical mode decomposition and local mean decomposition does not use the prior knowledge of the signal,and the basis function itself is obtained from the adaptive decomposition of the original signal,which overcomes the shortcomings of the integral time-frequency analysis method.Empirical mode decomposition is chosen as the research object of adaptive time-frequency analysis.In view of the characteristics of ground penetrating radar signals,the interpolation method in the decomposition process is discussed,and the spline interpolation method is determined to be suitable for the decomposition of GPR signals.The problem that the instantaneous frequencies of each component are inconsistent with the actual physical meanings in the initial and final periods is solved by the zero expansion method.Based on the layered geological model,two pollution models,shallow and deep,were designed respectively.Empirical modal decomposition of the time-distance profile of the pollution model was carried out to obtain the time-distance spectrum and frequency-distance spectrum of the corresponding different stages,and to explain the anomalous characteristics of the polluted area.In the field profile analysis,time-distance spectrum and frequency-distance spectrum effectively reflect the information of layered underground structure in the investigation area.According to the anomaly characteristics and drilling verification results,it is proved that the investigation area is not an industrial waste landfill site.Through the above research,the validity of empirical mode decomposition method in time-frequency analysis of ground penetrating radar signal is illustrated,and it can provide reliable auxiliary function for interpretation of ground penetrating radar profile.