Research On Topology Optimization Algorithm And Its Application In Antennas And Arrays

Topology optimization?TO?can produce holes or make holes disappear in the structure,which has a great design freedom and can provide schemes beyond the experience design.Moreover,the TO method based on gradient algorithm has fast convergence speed.The TO method based on gradient algorithm and its application in antennas and arrays are studied in this dissertation.A hybrid topology optimization?HTO?method combining material distribution method and level set method is used to design high-performance antennas and arrays.By improving the HTO method,the joint optimization of antenna matching,isolation and radiation characteristics is realized,the optimization design of patch antenna miniaturization,multiple-element antennas decoupling structures and wide-angle scanning phased arrays is carried out,and the excellent antenna performance is obtained.The main contents of this dissertation are as follows:Firstly,the material distribution method and the level set method,as well as the HTO method combining these two methods are studied and improved.The material distribution method has strong topology-searching ability,while the level set method can generate structures with smooth boundaries.The HTO method has the advantages of both methods,which can not only optimize satisfactory topology structures quickly,but also optimize structures with smooth boundaries.In order to optimize the more complex structures within smaller regions,a regularized Heaviside projection filter function is utilized to eliminate the blurry intermediate states.Due to the single objective function used in the current literature,the joint optimization of antenna S-parameters and radiation characteristics objectives is achieved by the adjoint sensitivity analysis method in this dissertation,so that the matching,isolation,gain and polarization objectives of an antenna can be optimized by the TO method based on gradient algorithm simultaneously.Besides,the application of symmetrical structure constraints to the HTO process reduces the number of optimization regions and optimization variables.Next,the design of antenna miniaturization is studied by using HTO method.Taking both the radiation characteristics and the matching as the optimization objectives,miniaturization of antenna patch is realized and good radiation performance is maintained simultaneously.Besides,the symmetrical radiation patch structure not only reduces the number of optimization regions and optimization variables,but also effectively suppresses the cross-polarization of the antenna without setting cross-polarization objectives.Finally,the radiation patch size of the optimized antenna is reduced by 63%,while the gain and efficiency of the antenna remain basically unchanged,reaching 5.9d Bi and 90%respectively.The number of iterations of antenna optimization is less than150,and the optimization time is within 7 hours.The correctness of the optimization results is verified by full-wave simulation and machining test.The optimized miniaturized antenna can be used as an element of the array to maintain good radiation performance at a small element distance.Then,the HTO method is adopted to study the decoupling structure optimization of multiple-element antennas.The optimization model is designed as a double-layer structure,with the top and bottom layers optimized simultaneously,and the mutual coupling is greatly reduced.Besides,multiple objectives such as matching,isolation,gain and polarization of the antenna are optimized comprehensively,not only the isolation is improved,but also good matching and radiation characteristics are effectively guaranteed.By optimizing the decoupling structures of the single-polarized MIMO antenna,the dual-polarized MIMO antenna and the multiple-element antenna,the isolation of the single-polarized MIMO antenna with a center distance of 0.25?₀,the dual-polarized MIMO antenna with a center distance of 0.35?₀ and the eight-element MIMO antenna with a center distance of 0.33?₀ is improved by 21 d B,6?11 d B and 10 d B respectively.The correctness of the optimization results is verified by full-wave simulation and test.At last,the HTO method is adopted to study the optimization design of dual-polarized wide-angle scanning phased array.Based on the magnetic dipole model,the element optimization model is designed,and a dual-polarized wide-beam antenna element with half-power beamwidth up to 166°is obtained by optimization.Nine elements are used to form a one-dimensional?1-D?phased array,achieving the 1-D scanning of±80°,with the gain fluctuation of less than 3 d B and the sidelobe level of less than-11.3 d B.The scanning range of the dual-polarized 1-D phased array designed by the HTO method is much better than that in the existing literature?±65°?,which proves that the HTO method has a strong ability to solve complex electromagnetic problems.