基于广义多项式混沌展开的无人机飞行性能不确定性分析

Uncertainty analysis of unmanned aerial vehicle flight performance using general polynomial chaos expansion

无人机飞行性能的不确定性分析有助于评估不确定性对无人机飞行任务规划的影响,为飞行任务方案的鲁棒性设计提供支撑。基于广义多项式混沌展开方法,建立了无人机飞行性能不确定性近似分析模型,利用样本实现了无人机性能多项式系数的非侵入式计算,研究了无人机状态在不确定条件下的非线性演化。在考虑初始条件、大气环境和气动参数存在不确定性的情况下,量化了不确定性对无人机性能的影响。数值分析结果表明,相比于传统的需要大量仿真样本的蒙特卡洛分析方法,提出的非侵入式展开方法在具有较高近似精度的情况下,仅需要少量的样本就可以完成无人机性能的不确定性分析,具有较高的计算效率。

Uncertainty analysis of unmanned aerial vehicles flight performance is useful for assessing the impact of uncertainties on unmanned aerial vehicles flight mission planning and supporting the robustness design of flight plans. An approximate analytical model of unmanned aerial vehicles flight performance under uncertainties was formulated based on the general polynomial chaos expansion method. The polynomial coefficients were calculated by nonintrusive method using samples. The nonlinear propagation of unmanned aerial vehicle states under uncertainties was analyzed. The influences of initial condition uncertainty, aerodynamic uncertainty and atmospheric uncertainty on vehicle performances were quantified. The results of numerical analysis show that comparing with the traditional Monte-Carlo analysis method which needs many simulation samples, the proposed nonintrusive expansion method with a high approximation precision only requires a relatively small quantity of samples, therefore has a high computational efficiency.