不确定条件下无人作战飞机在线攻击轨迹决策

摘要

Aimed at the uncertain factors in unmanned combat aerial vehicle (UCAV)air-to-surface com-bat,an online real-time attacking trajectory planning algorithm based on the receding horizon control (RHC) optimize strategy and Gauss pseudospectral method (GPM)is presented.The proposed algorithm adopts RHC strategy which transforms the global planning model into a series of overlapping but moving forward local opti-mize window,uncertain error can be eliminated through rolling update and feedback revise of this strategy, computational complexity and computational resource need can be greatly reduced through model predictive of this strategy;GPM is adopted to resolve a high-fidelity attacking trajectory which combines the flight envelop constraints,weapon delivery constraints and threat avoidance constraints;real-time iteration scheme is adopted to give initial Guess,and auto switch objective function is adopted to guide UCAV evade pop-up threat.Simula-tion result shows that the proposed algorithm can greatly eliminate uncertain error,and get a high-fidelity attac-king trajectory in real time.%针对无人作战飞机（unmanned combat aerial vehicle，UCAV）对地攻击过程中存在的诸多不确定性因素问题，提出基于滚动时域策略和 Gauss 伪谱法（Gauss pseudospectral method，GPM）的在线实时攻击轨迹决策算法。算法采用滚动时域的优化策略，将全局优化问题转化为一系列相互叠加但不断向前推进的优化区间，通过滚动更新和反馈校正消除不确定误差，同时降低了所提算法的计算复杂度和对计算资源的需求；综合飞行包线约束，武器投射约束和威胁规避约束，采用 GPM 完成高精度攻击轨迹求解；采用实时迭代策略提供优化初值，通过自适应目标函数切换引导 UCAV 迅速规避近距突发威胁。仿真结果表明，所提算法能够有效消除不确定误差，同时以较高的精度和速度生成可行的攻击轨迹。