In today’s evolving technology landscape, unmanned aerial vehicles (UAVs) have become a high-profile technology. Although reinforcement learning can successfully solve UAV path planning problems in simple environments, its research is still insufficient for complex tasks with time synchronization constraints. This paper primarily focuses on the rapid penetration strategy planning of UAV swarms against multiple targets. Aiming at the mission requirements of synchronized attacks by UAV swarms, a multi-target collaborative planning strategy for unmanned swarms based on the fusion of time constraints and migration reinforcement learning is proposed. This strategy adds time constraints on the basis of the swarm-to-single-target planning strategy, and achieves the simultaneous arrival of UAV swarms to multiple targets. The reward function of reinforcement learning is improved, and the training method of transfer reinforcement learning is adopted to improve the training efficiency.