Research On The Inversion Algorithm For Layered Media Parameters In Ground Penetrating Radar

Ground penetrating radar(GPR)has the advantages of convenient use,high detection efficiency and being suitable for various media.The technology of using GPR to obtain layered media information has been widely used in related fields such as highway quality inspection,municipal engineering,and environmental engineering.However,the existing parameter inversion algorithms have the problems of low precision,poor stability and poor real-time performance,and the results of waveform inversion cannot meet the practical needs.Although the two-way travel time of electromagnetic wave in the media layer can be used as prior information to improve the performance of the parameter inversion algorithm,it is still a difficult problem to obtain accurate two-way travel time.Therefore,this paper studies the inversion algorithm of layered media parameters and its performance from the following three aspects:1.An inversion method based on hybrid optimization algorithm is proposed to solve the problem of unstable and large error in the inversion results of GPR data in layered media.Firstly,the multi-level subdivision search method is used to determine the range of the relative permittivity parameters and reduce the search interval.Then,the particle swarm optimization(PSO)algorithm is used to search within the search interval to obtain the final result of the inversion.Reducing the search interval can exclude part of the local optimal solution,thus reducing the probability of particle swarm optimization falling into the local optimal solution.At the same time,it is also beneficial for the PSO algorithm to find the global optimal solution in the case of limited iteration times and particle number.Therefore,the accuracy and stability of the inversion are improved.The experimental results show that the hybrid optimization algorithm can improve the accuracy of the relative permittivity parameter and thickness of each layer of medium.The relative error and root mean square error are reduced,indicating that the stability of inversion results is increased.2.Although the above hybrid optimization algorithm has improved the stability and accuracy of the inversion results,it still cannot meet the needs of practical applications.Therefore,an inversion method based on resampling-particle swarm optimization(RS-PSO)algorithm is proposed to solve the problems of low precision,poor stability and poor real-time performance of the parameter inversion algorithm.In the RS-PSO algorithm,the proportional selection method is used for resampling.Through resampling,the particles with poor performance are discarded,and the particles with good performance are preserved and copied,which improves the optimization efficiency of the algorithm.The inertia factor is used to improve the iterative formula of the PSO algorithm,which improves the local search ability of the algorithm.The experimental results show that the proposed RS-PSO algorithm has greatly improved inversion accuracy,stability and real-time performance,and can effectively reconstruct the true structure of the relative permittivity parameters and thickness of the layered medium.3.Accurate two-way travel time is the key parameter of parameter inversion in layered media.The correlation-based layer picking algorithm tracks the seed points to pick the layer information by analyzing the correlation of echoes of adjacent traces.The two-way travel time of each layer can be obtained through the layer information.However,when the layer thickness changes greatly,the correct seed point cannot be found in the tracking process,which leads to the layer information error,resulting in a large error in the two-way travel time.Therefore,an improved correlation-based layer picking algorithm is proposed in this paper.The improved algorithm uses the first point with maximum amplitude outside the related windows to correct the seed point when locating a new seed point,which improves the accuracy of layer information and two-way travel time.The experimental results show that the accurate two-way travel time can be calculated by using the layer information obtained by the improved correlation-based layer picking algorithm.The improved algorithm combined with the RS-PSO algorithm can accurately invert the true structure of layered media.