复杂威胁环境下无人机实时航线规划逻辑架构

面向高对抗、强拒止的战场环境，实时航线规划是确保无人机（UAV）完成作战任务并提高自身生存概率的重要保障。为使无人机在面临不同程度的复杂威胁环境时能够选择合适的实时航线规划模式，提出了一种基于模糊推理机制的无人机实时航线规划逻辑架构。首先，对实时航线规划模式进行分类，从自主性的角度，重新划分人机权限分配等级，建立了实时航线规划模式与人机权限之间的联系；其次，针对典型观察—判断—决策—行动（OODA）循环存在“信任危机”的风险，构建了一种基于可变自主的实时航线规划体系架构，并对其逻辑进行了说明；最后，利用模糊推理机制实现了无人机系统动态人机权限分配，通过评判人机权限分配等级，进而确定实时航线规划模式。仿真结果表明:验证了实时航线规划逻辑架构的合理性和可变自主方法的有效性；经过综合分析，实时航线规划模式决策结果也比较符合实际作战需求；与模糊综合评价法相比，所提方法降低了人的主观性、实用性更强，得出的结果更加令人信服。 

UAV real-time route planning logical architecture in complex threat environment

Facing the battlefield environment with high confrontation and strong rejection, real-time route planning is an important guarantee to ensure the Unmanned Aerial Vehicle (UAV) to complete combat missions and improve its survival probability. In order to enable UAV to choose the appropriate real-time route planning mode when facing different levels of complex threat environment, a real-time route planning logic structure of UAV based on fuzzy inference mechanism is proposed. Firstly, the real-time route planning mode is classified. From the perspective of autonomy, the human-machine authority allocation levels are re-divided, and the connection between the real-time route planning mode and the human-machine authority is established. Secondly, aimed at the risk of "trust crisis" in typical Observation, Orientation, Decision, Action(OODA) control cycle, a real-time route planning architecture based on variable autonomy is constructed and its logic is explained. Finally, the dynamic human-machine authority allocation of UAV system is realized by using fuzzy inference mechanism, and the real-time route planning mode is determined by judging the man-machine permission assignment level. The simulation results show that the logic structure of real-time route planning is reasonable and the method of variable autonomy is effective. After comprehensive analysis, the decision-making results of real-time route planning mode also accord with the actual operational requirements. Compared with the fuzzy comprehensive evaluation method, the proposed method has lower subjectivity, stronger practicability and more convincing results.