Research On Synthetic Aperture Focusing Imaging Technique Of The Vehicle-mounted GPR Data

Vehicle-mounted GPR tunnel detection technique is a non-destructive,fast and continuous new method.It is the development direction of tunnel condition detection in the future.However,due to decrease of azimuth resolution and OCS interference in railway tunnel detection,the quality of GPR profile is affected.In this paper,a fast synthetic aperture focusing algorithm and a non-offset autocorrelation synthetic aperture focusing algorithm are proposed to make up for the reduction of azimuth resolution and to suppress OCS interference that are caused by the remote distance detection through OCS using vehicle-mounted GPR.The algorithm can improve the azimuth resolution of the vehicle mounted GPR and weaken the effect of the contact net on the effective signal.According to the characteristics and interference sources of the electrified railway tunnel detection by vehicle-mounted GPR,the adaptive synthetic aperture focusing imaging algorithm is proposed.Under the condition of improving the azimuth resolution,the strong interference target of the tunnel is focused to reduce the influence of interferon on the signal of the underground target.In order to connect other vehicle-mounted GPR data processing steps,the idea of "time-delay summation" is changed to "delayed summation and averaging" in time-domain.The vehicle-mounted GPR synthetic aperture focusing imaging process is derived in detail.In considering of huge amount of the vehicle-mounted GPR data,the package delay matrix "cell" is used to improve the operation speed.According to the characteristics of clutter interference in time-domain algorithm,a non-offset autocorrelation algorithm for synthetic aperture focusing algorithm is proposed.The two algorithms are applied to forward simulation data,and the speed and focusing ability of algorithms are analyzed.The fast synthetic aperture algorithm has the fastest computation speed,and non-offset autocorrelation synthetic aperture focusing algorithm has the best focusing effect.Both of them have increased the original peak energy by 159.83% and 356.41% respectively.The stability of algorithm is related to the accuracy of algorithm.The two algorithms in the relative permittivity error of ?3 range,the change is not more than 8%;the interface pick-up deviation is within the range of ?0.25 m,the change is not more than 15%;the two algorithms can adapt to the demand.Multi-target simulation is used to analyze the resolution of the algorithm in the case of multi-target interference.The results show that the two algorithms proposed in this paper can achieve good focusing effect and can distinguish different shapes and sizes of targets,and the non-offset autocorrelation synthetic aperture focusing algorithm has a better effect in multi-objective simulation.The effectiveness of the two algorithms are verified by using the Xicheng high speed railway tunnel data.The two algorithms can improve the azimuth resolution,focus on hyperbolas,improve the quality of the GPR profiles,and do not affect the interface reflections in the lining.The two algorithms can focus 5m width of the hyperbolia caused by the catenary column into a 0.5m wide image point on the GPR image,and the profile below the image point is not disturbed.For 1km tunnel data,the fast synthetic aperture focusing algorithm only needs about 10 min,and the non-offset autocorrelation synthetic aperture focusing algorithm needs about 60 min.These two algorithms can be used for pre-processing of tunnel detection data.The manual lifting antenna or hydraulic lifting antenna of GPR traditional detection method will meet the catenary column.It usually has to leak more than 3m before and after the catenary column.The vehicle-mounted GPR antenna is continuously detecting under the catenary,and there is no problem of leakage,but there is a hyperbolic interference reflected from the the catenary column in the radar image,and the two focusing imaging algorithm solves this problem well.This provides strong technical support for the application of vehicle-mounted GPR detection technology in high-speed railway tunnel detection.