基于Gauss伪谱法的高超声速飞行器多约束三维再入轨迹优化

高超声速飞行器再入飞行过程中,需要满足多种过程约束和终端状态约束,同时再入初始状态根据飞行任务不同会有较大变化,针对其特点的快速轨迹优化问题已成为当今热点。本文研究了一种基于"初值轨迹生成+Gauss伪谱法+SQP求解NLP"的方法,既利用了Gauss伪谱法收敛快、精度高的特点,又结合初值轨迹生成算法,弥补了Gauss伪谱法对初值敏感的不足。本文在仿真过程中选取再入总吸热量最小为性能指标,求解了满足多种约束的再入轨迹,并将优化的结果与数值积分的结果进行比较,验证了此算法有效性和可行性。

The Generation of Three-dimensional Constrained Entry Trajectories for Hypersonic Vehicle Based on the Gauss Pseudospectral Method

The multiple constraints are required to meet during the hypersonic vehicle reentry phase,including process constraints and terminal state constraints.Meanwhile the initial point of entry phase can be changed for different missions.Thus,the generation of trajectory for this kind of entry flights is focused nowadays.A method,making use of quick constringency and high precision of Gauss pseudospectral,called "Initial value creating + Gauss pseudospectral + SQP" is applied to compensate the deficiency of pseudospectral,which is sensitive to initial value.By choosing minimum decalescence as performance index,the three-dimensional constrained entry trajectories are generated,which is based on the advanced Gauss pseudospectral method stated above.The applicability and effectiveness of this method are demonstrated through the analysis of numerical simulation results.