协同空战与多智能体强化学习下的关键问题

自从协同作战的概念提出后,各军事强国在协同空战领域均取得了重大进展,协同成为提升作战能力的倍增器。近数十年来,作为解决序列问题的现代智能方法,强化学习在各领域高速发展。然而,面对高维变量问题时,传统的单智能体强化学习往往表现不佳,多智能体强化学习算法为解决复杂多维问题提出新的可能。通过对多智能体强化学习算法原理、训练范式与协同空战的适应性进行分析,提出了协同空战与多智能体强化学习的未来发展方向,为更好地把多智能体强化学习应用于协同空战提供思路。

Key Problems in Coordinated Air Combat andMulti-agent Reinforcement Learning

Since the concept of cooperative operation was put forward , all military powers have madegreat progress in the field of cooperative air combat , and coordination has become a multiplier to en.hance combat capability. In recent decades, as a modern intelligent method to solve sequence prob.lems , reinforcement learning has developed rapidly in various fields. However ,in the face of high-di.mensional variable problems ,the traditional single-agent reinforcement learning often performs poor.ly. Multi-agent reinforcement learning algorithms provide new possibilities for solving complex multi-dimensional problems. By analyzing the adaptability of multi-agent reinforcement learning algorithmprinciple , training paradigm and cooperative air combat, the future development direction of coopera-tive air combat and multi-agent reinforcement learning is proposed, which provides ideas for betterapplication of multi-agent reinforcement learning in cooperative air combat.