Research On Radar Imaging Technology Based On Bayesian Theory

For ideal radar detection environment,such as high SNR,no interference and stable radar platform movement,the traditional algorithm similar to range-Doppler algorithm can achieve good focusing imaging.However,for some complex radar operating environments and detection targets,the effectiveness of existing imaging methods will be reduced.Some complex environments can even produce colored noise,resulting in a lower SNR.In order to obtain good radar imaging results under complex conditions,this paper studies the radar imaging algorithm with high robustness based on Bayes theory under complex conditions.Most of the traditional radar imaging technologies acquire target images through range Doppler technology.The improvement of radar imaging performance will inevitably lead to the improvement of radar system complexity.In this paper,the geometric position relation between radar and target is established under the condition of static and relative motion of radar platform,and the echo signal model of scanning radar is deduced.In consideration of the wide application of radar correlation imaging method in improving the resolution of radar system,this paper establishes the radar correlation imaging echo model based on the research of radar echo model signal,and completes the system simulation.To solve the problem of poor imaging performance of existing algorithms at low SNR,the radar imaging method in complex environment based on bayesian theory is studied.In this paper,the traditional bayesian model is improved and an improved variable decibel bayesian radar imaging algorithm is proposed.The radar echo model is represented as a Gaussian mixture distribution bayesian model,and the imaginary part of the target vector is estimated by the variational approximation method based on the maximum posteriori,and the imaginary part of the target vector is estimated by the maximum likelihood estimation using the variational decibel method.This method respectively estimates the real and imaginary parts of the target,and turns the complex number domain of the bayesian prior model into the real number domain.In this paper,simulation experiments are carried out under different radar detection target scenarios.The simulation results show that the proposed algorithm improves the radar imaging performance and provides an effective method for highresolution radar imaging in complex environments.Aiming at the problem of radar model mismatch caused by various error factors in practical engineering application,this paper studies the radar imaging method under the model mismatch.The traditional correlation imaging radar needs to calculate the reference matrix accurately to reconstruct the image.For the model mismatch problem,this paper deduces the radar echo signal model under the model mismatch,and proposes a bayesian imaging algorithm based on joint least mean square estimation.In this algorithm,the lowest mean square error method is used to calculate the actual reference matrix,and the adaptive iterative updating variational estimation of the real reference matrix and radar detection target signals is realized by combining the calculation points of the VBMS.In addition,this paper USES the least square algorithm to estimate the interference matrix on the basis of the VBMR imaging algorithm,so as to correct the reference matrix with errors and realize the radar high-resolution imaging under the model mismatch.The simulation results of the above algorithms show that this paper has good imaging results,and has effective performance in noise suppression and adaptive model mismatch.