Research On Ground-based SAR Moving Target Detection Algorithm Based On Relative Speed

Mining slope stability and deformation information acquisition is an important information source for safety production,disaster prediction and early warning.Ground-based synthetic aperture radar(GBSAR)can obtain sub-millimeter deformation information of the observed area based on differential interference technology.This technology is an effective means for long-term and continuous monitoring.Ground-based SAR deformation inversion uses interferometric phase information.And the accuracy of deformation retrieval relies on high coherence of GBSAR image stack.However,the moving vehicles in the mining area will generate lots of defocused/displaced signatures on the SAR image,which can affect the interference phase and cause noise decoherence.Hence,such phenomenon reduces the accuracy of deformation inversion.One feasible way to eliminate the influence is to remove the refocused moving target.By detecting the moving target after refocusing,it can be removed from the original image.The existing GBSAR systems mostly use frequency modulated continuous wave(FMCW).At present,there is no research on the moving target detection of FMCW-GBSAR has been reported.Aiming at these problems above,this paper introduces the relative speed into the range equation,and deduces the moving target signal model based on the relative speed.Based on this model and the incomplete synthetic aperture of GBSAR,the Range Doppler(RD)algorithm is adopted and improved,and the moving target detection is completed by introducing the relative speed principle to achieve refocusing imaging.Specifically,the 2-dimensional parameters search space of relative speed and squint angle is established,and the 2D parameters are searched.The entropy value of the refocused image is calculated,and the moving target is refocused based on the principle of minimum entropy,so as to achieve the purpose of detecting the moving target.The main research contents of this paper are as follows:(1)The expression of FMCW-GBSAR moving target signal model is derived.Because this signal shows the complexity of developing the imaging algorithm,the relative speed is introduced into the expression.The introduction of relative speed effectively reduces the complexity of the signal model and lays the theoretical foundation for signal processing.Then,the expression of GBSAR moving target signal model based on relative speed is deduced.And the defocused and displaced characteristics of GBSAR moving target signals are analyzed.(2)According to the relative speed-based signal model,the Range Doppler(RD)algorithm is adopted and improved.And the relationship between radar velocity and moving target velocity is analyzed.The four detail conditions due to the fact that moving target speed faster than GBSAR are also presented,and the GBSAR moving target imaging algorithm based on relative speed refocusing is proposed.The expression of range cell migration correction phase multiplier at relative speed is deduced.Based on the incomplete synthetic aperture of GBSAR,a new method based on quadratic phase compensation of azimuth time domain signal is proposed to achieve azimuth pulse compression.Finally,the proposed algorithm is verified by simulation and actual data.(3)Aiming at the problem of moving target detection,the 2-dimensional parameters search space of relative speed and squint angle is established.And the moving target detection algorithm based on relative speed refocusing is proposed.The refocused target image can be obtained via searching 2D parameters.When the velocity parameter of the target matches the theoretical value,the moving target will be refocused.By calculating the entropy value of the moving target as the criterion of target detection,the purpose of detecting the moving target is achieved.Finally,the effectiveness of the algorithm is verified by simulation and actual data.