The Research On The Synthesis Techniques For Sparse Antenna Arrays

As a special part of the radar,the antenna design plays an important role in the radar system.Compared with the full filled array,the sparse array can not only reduce the size,weight and production cost of the system,but also keep the main beam width almost the same,improve the resolution,and have strong directivity.Therefore,the sparse array has been widely used.The peak sidelobe level(PSLL)is an important index to evaluate the performance of the array,while the PSLL will increase with the decrease of the number of array elements.In order to obtain a sparse array that meets the engineering needs,the element location and excitation need to be optimized.This paper discusses the optimal design of the position distribution and excitation of array elements under the condition of taking into account the array aperture,the number of array elements and the minimum array element spacing so as to obtain the lowest PSLL.In this paper,the basic theory of array antenna is introduced and the calculation formulas of the array pattern and the sparse distribution model of linear arrays,concentric ring arrays and rectangular grid plane arrays are given.The main study of the researches are linear arrays,concentric ring arrays and large scale of planar arrays.First of all,for the linear array,this paper uses a hybrid algorithm based on genetic algorithm and convex optimization algorithm to optimize the position and excitation of the sparse linear array.Besides,the adaptive mutation is used in the optimization process to generate new individuals,and the coding method is improved so that the searching space of the algorithm is reduced.Based on the optimized array element distribution,convex optimization algorithm is used to further optimize the excitation of the array,so that the PSLL of the pattern is significantly reduced.In the meantime,for the concentric circular arrays,this paper proposes a dimension reduction optimization technique based on continuous weighting density of reference circular aperture and recombination of margin code technique,and obtain constraints relationship between the number of elements per ring and the ring radius of the concentric ring arrays,The two-dimensional optimization problem is reduced to one-dimensional problem,which improves the search efficiency of the algorithm and greatly reduces the computational burden and complexity in the process of optimizing.Finally,the paper studies the problem of sparse optimization of large-scale planar arrays.By adopting the hybrid sparse optimization technology based on weighting density and genetic algorithm,the planar array is divided into several concentric ring rings,the number of array elements need to be turned on in each ring is obtained by weighting density.Besides by combining with the two-dimensional Fourier transform in the optimization process,the evaluation of the array pattern is speeding up and the optimization speed of the algorithm is improve.Simulation results verify the effectiveness of the above methods.