Study On Phase Only Pattern Synthesis For Large Array Antenna

The array pattern synthesis is one of the most important research directions in the field of array signal processing,which has important application significance.In order to obtain the expected array beam pattern,the excitation coefficient,the number and location of the array elements are optimized.The phase-only pattern synthesis only uses phase shifters to achieve beam control,which can achieve the consistency of array feed network and simplify its structure.This dissertation mainly studies the phase-only pattern synthesis algorithm for large array antennas.It focuses on the main lobe shaping,the reduction of side lobe level,and the requirements of multi-beam shaping.The main contents of this dissertation are as follows.Firstly,this dissertation introduces the basic theory of array pattern synthesis and establishes the signal model of planar array.Then,three classical pattern synthesis methods are listed,and some simulation experiments are given.The phase-only pattern synthesis model is established,in which only phase shifters is used to realize the pattern.Secondly,the alternating projection algorithm based on auxiliary phase optimization(APOAP)is introduced.Then,the excitation projection operator is improved.And the two-level excitation projection operator is given to limit the excitation amplitude quantitatively.Then,the phase only projection operator is given to achieve the consistency of array element excitation.The improved algorithm is applied to the pattern synthesis of large array antenna.Finally,hybrid genetic algorithm and particle swarm optimization(GA-PSO)algorithm is applied to the phase-only multi-beam pattern synthesis for low orbit satellites.The work area division scheme is designed,and the GA-PSO algorithm is applied to achieve phase-only multi beam shaping.Then,an improved invasive weed(PIWO)algorithm is proposed.In this algorithm,the Levy distribution is used as the global guidance.The mutation operation is introduced to avoid trapping into local optima.The competitive exclusion strategy is also improved to maintain the population diversity.Furthermore,in order to apply to phase-only pattern synthesis,an improved fitness function is given.The experimental results show that the improved algorithm has better synthesis performance.