Abstract: In order to improve the efficiency and accuracy of intelligent optimization algorithms in array antenna beam fouling, a multi-strategy improved artificial lemming algorithm (MSI-ALA) is proposed. The algorithm firstly classifies the selection of random individuals according to the number of iterations to balance the exploration and exploitation; secondly, it introduces the historical optimal solution and dynamic weights to the search radius to avoid over-reliance on the current local optimum; lastly, it proposes a new escape factor by integrating the fitness improvement rate collected in the iterative process to improve the optimization efficiency and robustness. The simulation and comparison tests of multiple intelligent optimization algorithms show that the improved algorithm has the advantages of fast convergence speed and high optimization accuracy. Applying the algorithm to the array antenna beam fouling optimization, the simulation results prove the applicability and efficiency of the improved algorithm, which is conducive to the further application of intelligent optimization algorithm and array antenna.