【摘要】 针对传统的火力分配算法仅适用于静态火力分配，无法估计战场态势的实时变化。文章提出多智能体近端策略优化方法，该方法可以处理分布式指挥下的动态火力分配问题。从实验结果可以看出，该方法能够考虑整个战场在时间维度上的相关性，相比于启发式方法，显著提升分配效果。对于分布式动态火力分配问题的求解有很好的借鉴意义。更多还原

【Abstract】 The traditional firepower allocation algorithm is only suitable for static firepower distribution and can not estimate the real-time change of battlefield situation. In this paper, a multi-agent near-end strategy optimization method is proposed, which can deal with the problem of dynamic firepower allocation under distributed command. From the experimental results, we can see that this method can consider the correlation of the whole battlefield in the time dimension, and significantly improve the allocation effect compared with the heuristic method. This paper has a good reference for the solution of distributed dynamic weapon target assignment.更多还原