Arrays Sidelobe Pattern Synthesis And Optimization

The array pattern synthesis algorithm is the main direction and difficulty in the research of phased array antenna.Array synthesis refers to control of the main lobe and sidelobe by using the position of array element distribution,the number of array elements,the excitation amplitude and the phase of array elements.In recent years,more restrictions and higher requirements are put forward for array synthesis with the expansion and development of application scenarios.It has become a hot research topic to apply intelligent optimization algorithms to array synthesis.In this thesis,we improve the Gravitational Search Algorithm(GSA)in the intelligent Algorithm and apply GSA to beam-shaping,sidelobe control and sparse array.The feasibility of the optimized results is verified in the practical array.The main contents of this thesis are as follows:(1)As an improved form of the GSA,the Chaotic Gravity Algorithm(CH-GSA)is proposed with the chaos operator.The CH-GSA is compared with the fundamental GSA and particle swarm optimization(PSO)algorithm by using phase-weighted super-cosecant-squared and flat-top beamforming.The global convergence and the improvement of convergence precision of the algorithm are verified,and the algorithm’s efficiency is improved.(2)At the same time,the GSA and Multi-Population Genetic Algorithm(MP-GA)are combined to get Multi-Population Gravity Algorithm(MP-GSA).MP-GSA is used in the phase-only weighted super-cosecant squares and sparse arrays,and compared with GSA and PSO algorithms.The main lobe fluctuates is 0.21 d B better than GSA and PSO in super-cosecant squares.The main lobe fluctuates are 0.28 d B obtained by GSA and 0.8d B obtained by PSO.At the same time,the left sidelobe satisfies the target value of-15 d B,and the right sidelobe satisfies the target value of-10 d B,while the GSA and the PSO algorithms can’t reach the target.The sidelobe level obtained by the MP-GSA algorithm is-18.07 d B in a sparse array,while the sidelobe level obtained by the GSA algorithm and PSO algorithm are worse than-18 d B,and the convergence rate is the fastest.The universal applicability and the improvement of convergence accuracy and speed are verified.(3)Finally,a printed dipole array antenna is designed as an array unit to form an array.The arrays’ elevation plane is shaped by a suppressant beam with phase weighting,and the phase data obtained by the MP-GSA algorithm is brought into the array.After processing and testing,the antenna index meets the requirements of main lobe beam-shaping and sidelobe.Secondly,a linear sparse array is composed of mixed feed-point elements,in which the MP-GSA algorithm is used to obtain the sparse array element position.The antenna array is processed and tested.The expected sidelobe is obtained.The practicability of the algorithm is verified by the processing test of two kinds of antenna arrays.