Research On Diagnosis And Optimization Of Phased Array Antenna Element Failure

Phased array antenna can control the excitation of array elements through T/R feed system,which can change the direction of radiation beam rapidly and shape beam.Thus,it is widely used in military radar,satellite communication,air traffic control and other fields.However,with the scale of the modern phased array antenna bigger and bigger,the array structure more and more complex,and long time working in the outdoor environment,it is extremely vulnerable to natural weather and other electromagnetic radiation interference,etc...Working after a period of time,some of elements of the phased array antenna would be failed.Therefore,it is very important to diagnose the position of the failed elements quickly,and then adjust the excitation of other elements when the failed elements cannot be replaced or repaired.The main contributions of this paper are as follows:(1)The effects of different failure rates and different failure modes on the pattern characteristics and beams of Taylor,Dolph-Chebyshev linear arrays,Chebyshev planar arrays and spherical arrays are analyzed.The simulation results show that with the increase of the failure rate,the peak side lobe level increases,but the main lobe width changes little.The closer the array element is to the center of the array,the higher the peak side lobe level is.The closer the failed element to the center of the arrays,the more the peak side lobe level rises.(2)When the diagnosis of the failed array elements,a new array is built with the difference between the failure-free array and the failed array to make it meet the requirements of the compressed sensing theory,thereby achieving the use of the far-field data that does not exceeding the number of array elements to reconstruct the excitation of the failed array elements,then quickly diagnose the location of the failed elements.When recovering the excitation of failed elements,select the appropriate ? and ? as the iteration parameters of TwIST algorithm,and compare this algorithm with the classic OMP algorithm,SAMP algorithm and CVX tool.In order to simulate the scene of actual sampled data,Gaussian noise is added into the diagnostic model.Through repeated independent experiments,these algorithms are used to diagnose the linear,planar and spherical array antennas with failed alements,and analyze the impact of different failure rates,sampling points and SNR on the diagnostic accuracy.(3)The alternating projection algorithm for synthesizing the array antenna pattern is applied to the optimization compensation of the remaining unfailed array elements.When the peak-to-average ratio does not exceed the peak-to-average ratio of original excitations,the main lobe width constraint is added to the target pattern of the alternating projection algorithm to optimize the peak side lobe level of the array after failure with far field sampling data whose number is twice the number of array elements.In the optimization model,the Gaussian noise is added to simulate the data sampled in the far field.First,under the condition of known the position of the failure elements,the alternative projection algorithm is used for optimization,which shows its effectiveness.Finally,TwIST algorithm is used to diagnose the location of the failure elements,and then according to the diagnosis result,the alternating projection algorithm is used to optimize the array after the failure.Through simulation experiments,the integration of diagnosis and optimization of the arrays with failure elements is realized.