Abstract: This study proposes an improved particle swarm optimization algorithm (PSO). Without modifying the original update formulas of standard PSO, the iterative update strategy derived from the grey wolf optimizer (GWO) is embedded into the PSO framework, which endows particles in the swarm with hierarchical population characteristics. By integrating the update iteration rule of the GWO into the standard PSO algorithm, the PSOGWO algorithm, a hybrid of PSO and GWO, was further developed to enhance convergence speed and accuracy for linear array antenna pattern beamforming. Additionally, for the first time, a learning factor asynchronous strategy is introduced into the PSOGWO to address issues related to local optimal solutions, resulting in an average improvement of 38.04% in convergence accuracy. Finally, simulations were conducted on array antennas using the proposed PSOGWO to validate its effectiveness in achieving cosecant square and flat-top patterns for beamforming.