Radar beamforming is one of the most important technologies in the radar system, and the anti unmanned aerial vehicle (anti-UAV) radar beamforming is usually difficult because of its complex beams. In this paper, a beamforming method based on a hybrid basic beams super position for anti-UAV radar and a modulation parameter decoupling method are proposed. Basic beams based on a window function and their improved versions with fast attenuation side lobe, are combined by different weightings on different pointing directions to obtain the anti-UAV radar beam. A basic beams generation method for improving the attenuation side lobe of the original basic beams is proposed by minimizing the total power without any power loss in one specific pointing direction. To obtain the optimal weightings for different direc tions, the average beamforming error is minimized and then solved by a genetic algorithm. The method for decoupling the modulation param eters is also derived mathematically. Three groups of experiments with different types of beams are carried out. The experimental results show that the proposed method with hybrid beams can obtain a better beam forming effect in terms of side lobe level based on the window function under different types of anti-UAV radar beams.