Research On Microwave Staring Correlated Sparse Imaging Of Spinning Target

Up to now the existing Microwave Staring Correlated Imaging(MSCI)algorithms are mainly focused on stationary targets or translational moving targets,which greatly limit the application of MSCI.In reality,many objects have complex motion patterns.Therefore,the research on microwave staring correlated imaging for complex moving targets is of great significance.This paper studies the spinning target under the existing research background to incorporate more flexible motion postures into the correlated imaging mechanism.The spinning target MSCI's modeling,analysis and the correlated imaging algorithms are also contained in this paper.Firstly,the off-grid problem in real scene is researched.We analyze the effect of off-grid problem for spinning target on imaging performance and propose an effective off-grid compensation algorithm.This paper uses the difference method to approximate the error of radiation field matrix caused by off-grid.Since the error is caused by the positions of the scatterer which are not located at the divided grid point,the scattering coefficients have the same sparse condition as the grid error.The approximate model can be derived as a block sparse problem and use the block sparse recovery method based on sparse Bayesian framework to solve.Secondly,correlated imaging algorithms for spinning are studied.In this paper,the effect of target's spinning is analyzed based on correlated imaging model.Under the sparse prior information of the target,the target's spinning speed and spinning center can be estimated roughly.Then based on the sparse Bayesian framework with the maximum posterior probability,more accurate spinning speed and imaging results can be obtained by two methods.One uses Taylor expansion to develop the nonlinear echo model with unknown spinning speed into a linear form and solves it by alternately updating spinning speed and scattering coefficients in each iteration step.Another transforms the problem into a block sparse problem and uses the block sparse algorithm to simultaneously get the spinning speed and scattering coefficients accurately.Finally,the MSCI method for the target with translation and spinning is studied.It is difficult to decouple the effect of the spinning center,spinning speed and the translation speed of the target to solve the imaging problem.In this paper,two different global search strategies based on adaptive grid search algorithm and particle swarm optimization algorithm are used.The former adaptively focuses the range of accurate parameter estimation on the parameters with better recovery results.When the quantity of the grids to be searched is very large,the FOCUSS algorithm can be used to accelerate the speed of the operation.When the grids need to be searched are fewer,SBL method is used for sparse recovery to improve the anti-noise performance of the algorithm.The latter randomly initializes a group of 3d particles by setting the appropriate fitness value related to the optimal solution and updates the individual extreme value,global extreme value and the current position of the particle's fitness value to obtain the global optimal solution in the parameters feasible regions.After iterations,we can get the estimated values of spinning center,spinning speed,translation speed and the corresponding radar image.