Research On Array Diagnosis Using Amplitude-Only Radiation Data

As the size of arrays increases,the probability of array element failure also increases.In actual applications,due to hardware or installation inanccuracy,arrays always suffer from array errors,namely array mismatch.These mismatches cause the actual array manifold differen from the ideal one.Therefore,array diagnosis that aims to locate the failing elements has been paid much attention.In this thesis,a research on failed sensor localization algorithm and array diagnosis against array mismatch using amplitude-only radiation field data is presented,specific tasks are follows:1.For array fault diagnosis using amplitude-only near field data,considering that most existing algorithms have specific application scenarios,a mixed norm failed sensor localization algorithm using amplitude-only near field data is presented,which is suitable for any array failure.Firstly,the prior information of failing probability of array elements is assumed to subject to the Laplacian distribution and the Gaussian distribution,with a parameter ? in fitting the actual situation.Then,the optimization problem based on the maximum posterior probability criterion is established.Finally,with alternating direction method of multipliers technique,the array excitation can be retrieved.In the simulation,it is shown that the proposed algorithm is more suitable for array fault diagnosis under arbitary failure rate than the existing algorithms,and it has better diagnostic results.2.Aiming at the problem of array mismatch in pratical applications,an iterative minimization algorithm using amplitude-only far field data is proposed.Firstly,the array model with array mismatch is appriximated by its first-order Taylor series expansion.The considered array mismatch include frequency shifting error and element position error.Then,the array mismatch and array excitation are sequentially fixed,and array excitation is retrieved through continuous iteration.In the simulation,it is shown that in the presence of array mismatch,the proposed algorithm gives more accurate array diagnosis results than the method that does not consider array mismatch.3.Based on the iterative minimization algorithm,an algotithm based on unknowns recombination using amplitude-only far field data is proposed.Unlike the iterative minimization algorithm,the array manifold is approximated by its first-order Taylor expansion.In particular,array mismatch and array excitation is linearly combined to a new variable,simplifying the original problem.Then,alternating direction method of multipliers is applied to solve the proposed optimization problem.It is shown that in the presence of array mismatch,the proposed method provides better accuracy of the restored array excitaion and small diagnosis error than the iterative minimization method and the method that does not consider array mismatch.