Research On Forward-looking Imaging Method Of Phased Array Radar

Airborne radar forward-looking imaging is widely used in military and civil fields,so the research on improving the resolution of radar forward-looking imaging is of great significance,which has become a hotspot of current research.Traditional imaging methods mainly use array or large aperture antenna to generate narrow beam for imaging,so the azimuthal resolution of radar is limited by real beam width and antenna aperture size,resulting in low resolution,difficult to achieve super resolution imaging,which is limited in practical application.Super-resolution imaging technology based on convolution inversion processes real beam scan data from the point of view of signal processing.The imaging method is simple and has a wide range of applications.It can achieve super resolution imaging by breaking through the limitation of antenna aperture.Therefore,based on the convolutional inversion technology,this paper studies the forward-looking imaging algorithm of phased array radar.The main work is as follows:(1)The forward scanning imaging signal model of phased array radar is deduced and established.The geometric relation between radar and target is established,and the expression of instantaneous distance between radar and target is obtained.Then the signal echo model is established by using the distance expression.In the case of radar movement,pulse compression and range-walking correction methods are used to preprocess the echo to eliminate the influence caused by platform movement.The signal echo can be written as the convolution result of antenna pattern and target radiation information,and the convolution model can be established.(2)The morbidity of the model was analyzed.When the antenna pattern and echo are known,the calculation of target scattering data can be regarded as an inverse process.Due to the existence of noise in the signal receiving process and the characteristics of antenna pattern matrix,the problem is ill-posed,the model becomes ill,and the direct inverse cannot get accurate results,so further algorithm research is needed.(3)Based on the regularization theory,an improved generalized total variation regularization algorithm is proposed for the ill-posed model.Considering the characteristics of radar signal data,real number constraints and amplitude-phase separation constraints are added to the model as regular terms,which make full use of the prior information of the target.The augmented Lagrange multiplier method is used to get the objective function,and the alternative direction multiplier method is used to solve the algorithm.At the same time,the effectiveness of the algorithm is analyzed by simulation experiment.(4)The model mismatch caused by the errors in the radar forward-looking imaging process is analyzed,and the influence of the errors on the solution results is deduced by using the least square method as an example,and verified by the simulation results.To solve the model mismatch problem,a generalized total variation regularization algorithm based on total least square estimation is studied.Firstly,the total least square method is used to estimate the error matrix,so as to realize the correction of the convolution matrix.Combined with the existing regularization methods,the high-resolution imaging under model mismatch is realized.Experimental results show that the proposed algorithm has good imaging performance.