GPR Instantaneous Attributes Analysis And Its Application In Near-Surface Survey

Ground penetrating radar(GPR)is a surveying method characterized by its convenience,speed,accuracy,and non-destructiveness.It obtains information about subsurface media by transmitting electromagnetic waves into the ground and receiving echoes reflected from below.Today,GPR is widely used in various engineering surveys.In the face of complex GPR acquisition profiles with many interferences,the processing and analysis of profile data is very important.The instantaneous attribute analysis of GPR data is a very effective time-frequency analysis method for GPR data.Whether it is instantaneous frequency,instantaneous amplitude,instantaneous phase or instantaneous centroid frequency,they have all performed well in the research of many scholars.This paper studies the instantaneous attribute analysis method of GPR data and applies it to the analysis of model simulation data and GPR field acquisition data.A comparative analysis of various instantaneous attributes of GPR obtained by different methods is carried out.The main research contents and results are as follows:In order to solve the ’mode mixing’ problem of Empirical Mode Decomposition(EMD),many sparse time-frequency decomposition methods of empirical mode have been proposed.In our research,we found that the Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(ICEEMDAN)performs best.It adaptively performs sparse time-frequency decomposition of GPR data and can avoid problems such as the ’end effect’ that plague EMD-like algorithms.On this basis,the extraction of instantaneous attributes for GPR data can better avoid noise interference.For the sparse time-frequency decomposition results(i.e.,Intrinsic Mode Functions,IMF),we introduce the Envelope Derivative Operator(EDO)as a new tool for obtaining GPR instantaneous attributes.Compared with the Hilbert transform,it can not only effectively ensure the non-negativity of instantaneous frequency when extracting instantaneous attributes,but also effectively suppress the ’end effect’ of the Hilbert transform.The three instantaneous attributes(instantaneous frequency,instantaneous amplitude,and instantaneous phase)of IMF slices obtained by ICEEMDAN-EDO can not only reflect the three instantaneous attribute information within different frequency bands but also avoid problems such as ’mode mixing’ and ’end effect’ that occur during the process of obtaining three instantaneous attributes based on Hilbert-Huang.The instantaneous centroid frequency of ground penetrating radar is a quantification of instantaneous amplitude in the frequency domain.It can effectively reduce the impact of low-frequency loss and has good tracking performance for underground targets.It has shown good robustness in GPR time-frequency analysis.Modeling and forward research were carried out on the collected research information of Chang’e series lunar penetrating radar(LPR),and the model simulation data was compared with the data collected by Chang’e 4 for instantaneous attribute analysis.It was proved that instantaneous attribute analysis is also applicable to radar near-surface exploration on the moon.