The dynamic sensor-weapon-target assignment (SWTA) problem in modern battlefield environments is characterized by high dynamism and strong adversariality. Traditional static assignment methods struggle to adapt to the rapidly evolving battlefield situation due to their low solving efficiency and inadequate environmental adaptability. To address these challenges, this paper proposes a dynamic SWTA method based on proximal policy optimization (PPO). By integrating the OODA (observe–orient–decide–act) loop theory, the method constructs realistic sensor detection probability and weapon kill probability models. Through continuous interaction between the agent and the environment, the PPO algorithm optimizes the strategy while balancing operational effectiveness and resource consumption. Experimental results demonstrate that the proposed method achieves superior performance across various ammo-to-target ratio scenarios, significantly enhancing both resource utilization and overall operational effectiveness. This study provides an efficient and adaptive intelligent decision-making framework for the dynamic SWTA problem, advancing the process of command and decision-making, with strong potential for practical application.