| Abstract: | Due to the shortcomings such as the premature convergence and the bad local optimal searching capability in traditional intelligence methods for pattern synthesis, a new type of wolf pack algorithm named Levy-Cultural Wolf Pack Algorithm (LCWPA) was designed on the basis of the Cultural Wolf Pack Algorithm (CWPA), which obeys the selective Levy flight. Because of the good overall management ability provided by the cultural algorithm in optimization process and the characteristics of excellent population diversity brought by Levy flight, the search efficiency of the new algorithm was greatly improved. When the algorithm was applied in the pattern synthesis of array antenna, the simulation results showed its high performance with multi-null and low side-lobe restrictions. In addition, the algorithm was superior to the Quantum Particle Swarm Optimization (QPSO), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA) in optimization accuracy and operation speed, and is of very good generalization. |
