Optimization Algorithms Study On The Beam-shaping Of Cylindrical Conformal Array Antennas

In this thesis, the pattern synthesis of a cylindrical conformal array antenna (CCAA) based on genetic algorithm (GA),particle swarm optimization (PSO) algorithm and their improvements is described.Firstly, the principle of GA and its realization methods are studied, a RGA is given; PSO algorithms is presented, the emphasis is played on the improvements methods for PSO. Two RGA's operations and a novel random-damp-reflection boundary condition are introduced into PSO, two new hybrid algorithms: Hybrid PSO (HPSO) and Parallel Hybrid PSO (PHPSO) are deduced. The four algorithms are applied in pattern synthesis of a liner array antenna, and their validities in this application are proved. The algorithms are characterized by the random statistic method, which is a useful attempt to describe the algorithms'performance; the results show that HPSO and PHPSO have better performance in the convergence ability and stability.Secondly, a CCAA model with directional radiating elements is established. The four algorithms are adopted to synthesize the pattern of the CCAA; they are proved to be feasible in this application, the statistic results show that HPSO is the steadiest one in convergence. HPSO is used to synthesize the pattern of CCAA of different parameters; the simulated patterns indicate that if the parameters are selected properly, then satisfied results can be obtained.Finally, the optimization of CCAA is summarized, and some useful conclusions are obtained.