多无人机协同空战任务规划仿真系统

The multi-UAV cooperative air combat task planning simulation system

面向多无人机协同空战任务规划方法性能验证的需求,开发了一种基于VR-Forces仿真引擎的多无人机协同空战任务规划仿真系统,对该仿真系统的总体架构、红方子系统、蓝方子系统、白方子系统和强化学习算法训练子系统进行了设计。对算法库、强化学习算法训练和人机交互等关键技术提出了针对性的解决方案,提升了该仿真系统的稳定性、可扩展性和功能性。利用该仿真系统对多无人机协同空战目标分配、重决策算法等典型任务规划方法进行了测试和仿真实验。结果表明,所开发的仿真系统可以很好地支持多无人机协同空战任务规划方法的测试和验证。

To verify the performance of multi-UAV cooperative air combat task planning method,a multi-UAV cooperative air combat task planning simulation system based on the VR-Forces simulation engine is developed in this paper.The architecture,red-side subsystem,blue-side subsystem,white-side subsystem,and reinforcement learning algorithm training subsystem of the simulation system are designed.To improve the stability,scalability,and functionality of the simulation system,key technologies such as algorithm library,reinforcement learning algorithm training,and human-computer interaction are proposed.On this basis,the simulation system is developed.Finally,the developed simulation system is used to test and verify the typical task planning methods,such as multi-UAV cooperative air combat target allocation and the redecision algorithm.The results show that the developed simulation system can support the testing and verification of the multi-UAV cooperative air combat task planning method.