基于深度强化学习的智能作战决策研究

Research on intelligent combat decision making based on deep reinforcement learning

为实现真正意义上的智能化、自主化无人作战,提出基于AI开发板构建作战任务计算机作为无人车的控制核心,模拟无人车作战机动态势图,并利用深度强化学习网络DQN建立角度及距离决策网络,实现无人车的智能机动决策。通过实验证实了无人车能够自主机动至目标区域,通过深度强化学习网络能实现平台自主、智能化决策,可为构建作战任务计算机实现真正意义上的智能化、自主化、无人化作战,提供可行技术途径及理论支撑。

With the appearance of the advantages of an unmanned combat platform in modern wars,research on unmanned combat platforms has become the focus of all circles.In order to realize intelligent and autonomous unmanned combats in a real sense,this paper constructs a combat mission computer as the control core of unmanned vehicles based on AI development board.It also simulates unmanned vehicle combat mobility situation diagram,and realizes intelligent maneuvering decisions of unmanned vehicles by using the deep reinforcement learning network DQN to establish the angle and the distance decision network.The experiment verifies that the unmanned vehicle can maneuver to the target area autonomously.Through the deep reinforcement learning network,the platform realizes autonomous and intelligent decision making,which provides a feasible technical approach and theoretical support for the construction of combat mission computers to realize intelligent,autonomous and unmanned combats.